Innodata Inc. is a data engineering and AI systems services company that supports the development, training, post-training, evaluation, and deployment of advanced artificial intelligence systems. The company provides training and post-training data development, alignment and preference optimization, capabilities and safety evaluation, and AI enablement services, including support for agentic and tool-using systems. Innodata serves technology companies, frontier AI laboratories, and enterprises, with clients including five of the "Magnificent Seven" tech companies. The company offers AI-enabled industry platforms, including Synodex for transforming medical records into structured digital data for insurance and healthcare workflows, and Agility PR Solutions for media intelligence. Founded in the late 1980s, Innodata began building proprietary AI language models in 2016-2017 and has since developed integrated AI lifecycle data solutions. The company conducts red teaming, adversarial evaluations, and safety research, including threat modeling for agent-based systems and assessment of data exfiltration risks. Innodata also provides AI consulting, strategy, and governance services, as well as solutions for U.S. government missions with security-cleared expertise.

Summarised from the company's own disclosures, fiscal year ended Feb 2026

Industry Information Technology Services Homepage innodata.com
Sector Technology IR page investor.innodata.com/overview/default.aspx
Tags
ai data engineering, model evaluation, red teaming, human preference optimization, multimodal data collection, agentic ai, ai safety, reinforcement learning environments, computer vision data, ai governance, Synodex, Agility PR Solutions
Press page —

Metrics

Valuation  
Market cap $ 2.263B
Market cap (USD) $ 2.263B
P/E 47.65
PEG —
P/BV 12.82
P/S 6.46
P/OCF 24.24
P/FCF 27.32
Dividend yield —
Turnover / mcap —
Returns  
ROE 37.7%
ROI 114.3%
ROIC 110.1%
Profit margin 12.8%
Growth  
Revenue growing Yes
Profits growing Yes
Revenue up 3y Yes
Revenue up 5y Yes
Earnings growth 3y (net profit CAGR) —
Quality  
Profitable Yes
Margin expanding Yes
Buying back stock No
Diluting (TTM) Yes
Debt/Equity —
Podcast bulls (TTM) Sign up
Podcast bears (TTM) Sign up
Podcast sentiment (TTM) Sign up
Podcast sentiment % (TTM) Sign up
News sentiment (TTM) 35
Quality score 33.3%
Insiders  
Buys, 1w Sign up
Buys, 1m Sign up
Buys, 1y Sign up
Sells, 1w Sign up
Sells, 1m Sign up
Sells, 1y Sign up

Sign up to see this company's insider trades

Figure FY 2025 FY 2024 FY 2023 FY 2022 FY 2021 FY 2020 FY 2019 FY 2018 FY 2017 FY 2016 FY 2015 FY 2014 FY 2013 FY 2012 FY 2011 FY 2010 FY 2009 FY 2008 FY 2007 FY 2006 FY 2005 FY 2004 FY 2003 FY 2002 FY 2001 FY 2000 FY 1999
Revenue 251.663M 170.461M 86.775M 79.001M 69.755M 58.24M 55.858M 57.418M 60.929M 63.074M 58.523M 59.076M 64.249M 86.591M 73.942M 61.513M 79.329M 75.001M 67.731M 40.953M 42.052M 53.949M 36.714M 36.385M 58.278M 50.731M 27.49M
Net profit 32.181M 28.66M -908K -11.935M -1.673M 617K -1.602M 4K -5.055M -5.524M -2.826M -974K -10.632M 7.473M 4.471M -747K 7.313M 7.584M 4.566M -7.323M -1.651M 7.857M 475K -5.165M 1.348M 6.168M 2.113M
Free cash flow 35.648M 27.274M 339K -7.742M 783K 4.246M 3.084M 1.571M -2.672M -5.475M 1.962M 2.58M -3.159M 10.92M -8.866M 2.986M 9.927M 2.052M 1.548M -5.783M -1.199M 13.618M -1.726M 1.888M -728K 4.984M -1.013M

Figure FY 2025 FY 2024 FY 2023 FY 2022 FY 2021 FY 2020 FY 2019 FY 2018 FY 2017 FY 2016 FY 2015 FY 2014 FY 2013 FY 2012 FY 2011 FY 2010 FY 2009 FY 2008 FY 2007 FY 2006 FY 2005 FY 2004 FY 2003 FY 2002 FY 2001 FY 2000 FY 1999
Share price 50.95 39.52 8.14 2.97 5.92 5.3 1.14 1.5 1.36 2.45 2.85 2.92 2.45 4.29 — 2.7 6.7 3.03 — 1.84 3.5 — — — — — —
Price performance 12m 0.29 3.86 1.75 -0.5 0.12 3.65 -0.24 0.1 -0.45 -0.14 -0.02 0.19 -0.43 — — -0.6 1.21 — — -0.48 — — — — — — —

Filed figures

Figure Dec 2016 Dec 2017 Dec 2018 Dec 2019 Dec 2020 Dec 2021 Dec 2022 Dec 2023 Dec 2024 Dec 2025
Balance sheet detail
Trade receivables (USD) 9.95M 10.3M 10.6M 9.72M 10M 11.4M 9.53M 14.3M 28M 46.5M
↳ YoY growth +7.6% +3.4% +3.3% -8.5% +3.3% +13.2% -16.3% +50.0% +96.1% +66.0%
↳ 5y CAGR -14.4% -6.4% -2.2% -1.4% +1.7% +2.7% -1.5% +6.1% +23.6% +35.9%
Borrowings, current (USD) 1.12M 2.13M 1.53M 912,000 1.71M 1.28M 877,000 — — —
Borrowings, non-current (USD) 3.92M 4.48M 4.06M 4.53M 6.28M 6.22M 5.08M — — —
Lease liabilities (USD) — — 0.00 7.84M 7.32M 6.31M 4.73M 5.48M 4.66M 4.43M
Cash and deposits (USD) 14.2M 11.4M 10.9M 10.9M 17.6M 18.9M 9.79M 13.8M 46.9M 82.2M
Intangible assets (USD) 8.19M 7.61M 6.28M 5.48M 4.66M 10.3M 12.5M 13.8M 13.4M 14M
Right-of-use assets (USD) — — 0.00 7.01M 6.61M 5.62M 4.31M 5.05M 4.24M 4.09M
Net debt (borrowings + leases - cash) (USD) -9.14M -4.8M -5.28M 2.41M -2.26M -5.1M 893,000 — — —
Cash flow detail
Operating cash flow (USD) -2.74M 738,000 3.6M 4.85M 5.66M 5.15M -1.22M 5.9M 35M 46.8M
Purchases of property, plant and equipment (USD) -2.74M -3.41M -2.03M -1.77M -1.41M -4.37M -6.53M -5.56M -7.74M -11.1M
Change in receivables (cash effect) (USD) 17,000 -40,000 -533,000 1.22M -481,000 -1.87M 1.3M -5.12M -14.7M -18.4M
Change in payables (cash effect) (USD) -1.75M 1.6M -779,000 -589,000 155,000 4.44M 322,000 -490,000 3.43M 9.5M
Share buybacks (cash effect) (USD) -134,000 0.00 — — — — — — — —
Interest paid (cash effect) (USD) — — — -130,000 -141,000 -28,000 -19,000 -400,000 -287,000 -229,000
Income taxes paid (cash effect) (USD) -1.31M -1.09M -680,000 -962,000 -348,000 -1.54M -1.11M -753,000 -2.42M -6M
Share-based compensation (USD) 1.16M 695,000 796,000 836,000 913,000 1.75M 3.28M 4.03M 4M 11.1M
Operating cash flow before change in receivables (USD) -2.75M 778,000 4.14M 3.64M 6.14M 7.02M -2.52M 11M 49.7M 65.1M
Free cash flow (OCF - PPE - intangibles) (USD) -5.48M -2.67M 1.57M 3.08M 4.25M 783,000 -7.74M 339,000 27.3M 35.6M
Income detail
Total comprehensive income (USD) -6.15M -4.19M -850,000 -2.52M 626,000 -3.06M -11.9M -402,000 27.8M 32.5M
Comprehensive income to owners (USD) -5.76M -3.89M -857,000 -2.51M 599,000 -2.93M -11.9M -421,000 27.8M 32.5M
Provision for credit losses (USD) 136,000 -139,000 0.00 — — — — 426,000 804,000 108,000
Credit cost ratio (credit losses / average receivables) (%) 1.4% 1.4% 0.0% — — — — 3.6% 3.8% 0.3%
Reported Mar 15, 2017 Mar 22, 2018 May 10, 2019 Mar 15, 2021 Mar 15, 2021 Mar 24, 2022 Feb 24, 2023 Mar 4, 2024 Feb 24, 2025 Feb 26, 2026

Sign up to see institutional holdings

Sign Up

Podcast Mentions

Sign up to see podcast mentions and sentiment

Sign Up

Innodata Reports Record Third Quarter 2026 Results Bull 5 Bear 0 Results

11/05/2026

  • Revenue Up 53% Year-Over-Year, Beats Consensus by 15%
  • Adjusted EBITDA of $17.7 Million, Beats Consensus by 15%
  • Raises Full-Year 2026 Revenue Growth Guidance to ~42% or More, Up From ~40% or More

NEW YORK, NY / ACCESS Newswire / November 5, 2026 / INNODATA INC. (Nasdaq:INOD) today reported results for the third quarter ended September 30, 2026.

  • Revenue of $95.6 million, representing 53% year-over-year revenue growth.
  • Adjusted Gross Profit of $38.3 million, representing Adjusted Gross Margin of 40%.*
  • Adjusted EBITDA of $17.7 million, or 18.5% of revenue, an increase of $1.5 million from $16.2 million in the same period last year.*
  • Net income of $10.3 million, or $0.30 per basic share and $0.29 per diluted share for the three-month period ended September 30, 2026, compared to net income of $8.3 million, or $0.26 per basic share and $0.24 per diluted share, in the same period last year.
  • Cash and cash equivalents of $164.0 million as of September 30, 2026, an increase of $81.7 million from $82.2 million as of December 31, 2025. Cash as of September 30, 2026 includes customer prepayments related to pass-through costs; net of these prepayments, cash was approximately $153.7 million as of September 30, 2026.

* Adjusted Gross Profit, Adjusted Gross Margin, and Adjusted EBITDA are non-GAAP financial measures and are defined below.

Rahul Singhal, CEO, said, "Q3 was another record revenue quarter for Innodata. Revenue of $95.6 million grew 53% year-over-year and exceeded analyst consensus by approximately $12.3 million, or 15%, marking our 13th consecutive quarter of year-over-year growth. Adjusted EBITDA of $17.7 million exceeded consensus by approximately 15%, net income of $10.3 million grew 24% year-over-year, and fully diluted earnings per share of $0.29 beat consensus by approximately 70%. This is my first earnings report as CEO, and I am pleased to be starting with results like these.

"The story of the quarter is the breadth of what we are winning. A customer that represented less than 5% of our revenue a year ago became our largest customer in Q3, giving us two customers of comparable scale and meaningfully advancing our diversification. We have already signed an additional program with this customer in Q4 valued at approximately $18 million. Separately, we won five new programs with a large hyperscaler from which we generated approximately $3.1 million of revenue in the trailing 12 months; based on those wins, we are now forecasting more than $20 million of revenue from this customer in 2027. We also added five new logos in Q3, including model builders and an enterprise data platform.

"We believe we are on the cusp of a profound market opportunity in physical AI. We understand that capital is flowing into robotics and embodied intelligence faster than into any category since the LLM boom began, and robot foundation models improve with data the same way LLMs do. We launched our motion capture lab in September for humanoid robot and world model training, and we are in pilots with leading frontier and robotics labs. We anticipate just one of these pilots, if successful, will yield approximately $7 million in revenue in 2027.

"We also continue to gain traction beyond the frontier labs. In enterprise AI, we added new relationships with a global technology hardware company and a major financial services firm. In Federal, we have a verbal win for a contract valued at $5 million per year for five years, which we expect to sign by the end of the fourth quarter; it would be the third program our Federal practice has won through our prime partners in its first year.

"With this momentum, we are raising our full-year 2026 revenue growth guidance to 42% or more year-over-year, up from 40% or more, even after absorbing the wind-down of a program at one of our large customers that we expect to be replaced by new programs in Q1. We believe we are positioned at the center of the most important transition in AI - the move to agents, reinforcement learning and embodied intelligence - and of the trust and safety work that we believe will be required to make the transition succeed. These are, at their core, data challenges, and they are squarely in our wheelhouse. I believe 2026 will finish as a strong year for Innodata, and that 2027 will prove even more exciting. It is a privilege to be leading this company at this moment."

Amounts in this press release have been rounded. All percentages have been calculated using unrounded amounts.

Timing of Conference Call with Q&A

Innodata will conduct an earnings conference call, including a question-and-answer period, at 5:00 PM eastern time today. You can participate in this call by dialing the following call-in numbers:

The call-in numbers for the conference call are:

(+1) 833 461 5787 North America-Toll Free
(+1) 585 542 9983 International
Participant Access Code 730 768 717

For Replay:

(+1) 833 309 1852 North America-Toll Free
(+44) 800 007 4044 United Kingdom
(+1) 929 828 5978 International
Access Code 730 768 717

It is recommended that participants dial in approximately 10 minutes prior to the start of the call. Investors are also invited to access a live Webcast of the conference call at the Investor Relations section of Innodata's website at https://investor.innodata.com/events-and-presentations/ . Please note that the Webcast feature will be in listen-only mode.

Call-in replay will be available for seven days following the conference call, and Webcast replay will be available for 30 days following the conference call, at the Investor Relations section of Innodata's website at https://investor.innodata.com/events-and-presentations/ .

About Innodata

Innodata (Nasdaq: INOD) is a global data engineering company. We believe that data and Artificial Intelligence (AI) are inextricably linked. Our mission is to enable the responsible advancement of artificial intelligence by providing the data, evaluation frameworks, and human expertise required to build AI systems that can be trusted at scale. We provide a range of transferable solutions, platforms, and services for Generative AI / AI builders and adopters. In every relationship, we honor our 36+ year legacy delivering the highest quality data and outstanding outcomes for our customers.

Visit www.innodata.com to learn more.

Forward-Looking Statements

This press release may contain certain forward-looking statements within the meaning of Section 21E of the Securities Exchange Act of 1934, as amended, and Section 27A of the Securities Act of 1933, as amended. These forward-looking statements include, without limitation, statements concerning our operations, economic performance, financial condition, developmental program expansion and position in the AI services market. Words such as "project," "forecast," "believe," "expect," "can," "continue," "could," "intend," "may," "should," "will," "anticipate," "indicate," "guide," "predict," "likely," "estimate," "plan," "potential," "possible," "promises," or the negatives thereof, and other similar expressions generally identify forward-looking statements.

These forward-looking statements are based on management's current expectations, assumptions and estimates and are subject to a number of risks and uncertainties, including, without limitation, impacts resulting from ongoing geopolitical conflicts; anticipated and actual use cases and outcomes; investments in large language models; that contracts may be terminated by customers; projected or committed volumes of work may not materialize; pipeline opportunities and customer discussions which may not materialize into work or expected volumes of work; the likelihood of continued development of the AI markets, particularly new and emerging markets, that our services support; the ability and willingness of our customers and prospective customers to execute business plans that give rise to requirements for our services; continuing reliance on project-based work and the primarily at-will nature of such contracts and the ability of these customers to reduce, delay or cancel projects; potential inability to replace projects that are completed, canceled or reduced; revenue concentration among a limited number of customers; our dependency on third-party providers and partners; our ability to achieve revenue and growth targets; difficulty in integrating and deriving synergies from acquisitions, joint ventures and strategic investments; potential undiscovered liabilities of companies and businesses that we may acquire; potential impairment of the carrying value of goodwill and other acquired intangible assets of companies and businesses that we acquire; a continued downturn in or depressed market conditions; changes in external market factors; the potential effects of U.S. global trade and monetary policy, including the interest rate policies of the Federal Reserve; changes in our business or growth strategy; the emergence of new, or growth in existing competitors; various other competitive and technological factors; our use of and reliance on information technology systems, including potential security breaches, cyber-attacks, privacy breaches or data breaches that result in the unauthorized disclosure of consumer, customer, employee or company information, or service interruptions; and other risks and uncertainties indicated from time to time in our filings with the Securities and Exchange Commission ("SEC").

Our actual results could differ materially from the results referred to in any forward-looking statements. Factors that could cause or contribute to such differences include, but are not limited to, the risks discussed in Part I, Item 1A. "Risk Factors," Part II, Item 7. "Management's Discussion and Analysis of Financial Condition and Results of Operations," and other parts of our Annual Report on Form 10-K, filed with the SEC on February 26, 2026, and in our other filings that we may make with the SEC. In light of these risks and uncertainties, there can be no assurance that the results referred to in any forward-looking statements will occur, and you should not place undue reliance on these forward-looking statements. These forward-looking statements speak only as of the date hereof.

We undertake no obligation to update or review any guidance or other forward-looking statements, whether as a result of new information, future developments or otherwise, except as may be required by the U.S. federal securities laws.

Company Contact

Aneesh Pendharkar
[email protected]

(201) 371-8000

Non-GAAP Financial Measures

In addition to the financial information prepared in conformity with U.S. GAAP ("GAAP"), we provide certain non-GAAP financial information. We believe that these non-GAAP financial measures assist investors in making comparisons of period-to-period operating results. In some respects, management believes non-GAAP financial measures are more indicative of our ongoing core operating performance than their GAAP equivalents by making adjustments that management believes are reflective of the ongoing performance of the business.

We believe that the presentation of this non-GAAP financial information provides investors a more complete understanding of our financial performance, competitive position, and prospects for the future, particularly by providing the same information that management and our Board of Directors use to evaluate our performance and manage the business. However, the non-GAAP financial measures presented in this press release have certain limitations in that they do not reflect all of the costs associated with the operations of our business as determined in accordance with GAAP. Therefore, investors should consider non-GAAP financial measures in addition to, and not as a substitute for, or as superior to, measures of financial performance prepared in accordance with GAAP. Further, the non-GAAP financial measures that we present may differ from similar non-GAAP financial measures used by other companies.

Adjusted Gross Profit and Adjusted Gross Margin

We define Adjusted Gross Profit as revenues less direct operating costs attributable to Innodata Inc. and its subsidiaries in accordance with GAAP, plus depreciation and amortization of intangible assets, stock-based compensation and other one-time costs included within direct operating cost.

We define Adjusted Gross Margin by dividing Adjusted Gross Profit over total GAAP revenues.

We use Adjusted Gross Profit and Adjusted Gross Margin to evaluate results of operations and trends between fiscal periods and believe that these measures are important components of our internal performance measurement process.

A reconciliation of Adjusted Gross Profit and Adjusted Gross Margin to the most directly comparable GAAP measure is included in the tables that accompany this release.

Adjusted EBITDA

We define Adjusted EBITDA as net income attributable to Innodata Inc. and its subsidiaries in accordance with GAAP before interest expense, income taxes, depreciation and amortization of intangible assets (which derives EBITDA), plus additional adjustments for loss on impairment of intangible assets and goodwill, stock-based compensation, income attributable to non-controlling interests and other one-time costs.

We use Adjusted EBITDA to evaluate core results of operations and trends between fiscal periods and believe that this measure is an important component of our internal performance measurement process.

A reconciliation of Adjusted EBITDA to the most directly comparable GAAP measure is included in the tables that accompany this release.

INNODATA INC. AND SUBSIDIARIES
CONDENSED CONSOLIDATED STATEMENTS OF OPERATIONS
(Unaudited)
(In thousands, except per-share amounts)

Three Months Ended Nine Months Ended
September 30, September 30,
2026 2025 2026 2025
Revenues $ 95,618 $ 62,550 $ 277,856 $ 179,287
Operating costs and expenses:
Direct operating costs 60,081 37,046 160,067 107,508
Selling and administrative expenses 26,882 13,745 76,374 42,837
Interest income, net (1,334 ) (420 ) (3,471 ) (1,124 )
85,629 50,371 232,970 149,221
Income before provision for income taxes 9,989 12,179 44,886 30,066
Provision for income taxes (340 ) 3,837 5,247 6,718
Consolidated net income 10,329 8,342 39,639 23,348
Income attributable to non-controlling interests - - - -
Net income attributable to Innodata Inc. and Subsidiaries $ 10,329 $ 8,342 $ 39,639 $ 23,348
Income per share attributable to Innodata Inc. and Subsidiaries:
Basic $ 0.30 $ 0.26 $ 1.18 $ 0.74
Diluted $ 0.29 $ 0.24 $ 1.13 $ 0.67
Weighted average shares outstanding:
Basic 34,389 31,848 33,497 31,690
Diluted 35,622 35,266 35,119 34,996

INNODATA INC. AND SUBSIDIARIES
CONDENSED CONSOLIDATED BALANCE SHEETS
(Unaudited)
(In thousands)

September 30, 2026 December 31, 2025
ASSETS
Current assets:
Cash and cash equivalents $ 163,967 $ 82,230
Accounts receivable, net 44,399 46,510
Prepaid expenses and other current assets 24,730 6,654
Total current assets 233,096 135,394
Property and equipment, net 8,367 7,966
Right-of-use asset, net 3,281 4,094
Other assets 3,002 1,648
Deferred income taxes, net 5,404 3,429
Intangibles, net 14,356 13,983
Goodwill 2,043 2,079
Total assets $ 269,549 $ 168,593
LIABILITIES AND STOCKHOLDERS' EQUITY
Current liabilities:
Accounts payable $ 6,156 $ 9,615
Advances from customers 8,607 1,812
Accrued expenses and other liabilities 32,999 7,800
Accrued salaries, wages and related benefits 17,567 16,480
Deferred revenue 7,887 7,493
Income and other taxes 5,045 4,471
Long-term obligations - current portion 2,370 1,659
Operating lease liability - current portion 1,333 1,202
Total current liabilities 81,964 50,532
Deferred income taxes, net 50 146
Long-term obligations, net of current portion 8,890 7,625
Operating lease liability, net of current portion 2,195 3,228
Total liabilities 93,099 61,531
STOCKHOLDERS' EQUITY 176,450 107,062
Total liabilities and stockholders' equity $ 269,549 $ 168,593

INNODATA INC. AND SUBSIDIARIES
CONDENSED CONSOLIDATED STATEMENTS OF CASH FLOWS
(Unaudited)
(In thousands)

Nine Months Ended
September 30,
2026 2025
Cash flows from operating activities:
Consolidated net income $ 39,639 $ 23,348
Adjustments to reconcile consolidated net income to net cash
provided by operating activities:
Stock-based compensation 19,918 8,309
Depreciation and amortization 6,730 4,913
Deferred income taxes, net (2,029 ) 3,201
Pension cost 1,181 984
Changes in operating assets and liabilities:
Accounts receivable 1,945 (11,167 )
Prepaid expenses and other current assets (17,284 ) (299 )
Other assets 421 (248 )
Accounts payable, accrued expenses and other 28,546 5,963
Accrued salaries, wages and related benefits 1,103 1,397
Income and other taxes 603 (2,530 )
Pension benefit payments (289 ) (139 )
Net cash provided by operating activities 80,484 33,732
Cash flows from investing activities:
Capital expenditures (7,714 ) (8,286 )
Net cash used in investing activities (7,714 ) (8,286 )
Cash flows from financing activities:
Proceeds from exercise of stock options 10,952 1,548
Withholding taxes on net settlement of restricted stock units (319 ) (37 )
Payment of long-term obligations (849 ) (230 )
Net cash provided by financing activities 9,784 1,281
Effect of exchange rate changes on cash and cash equivalents (817 ) 235
Net increase in cash and cash equivalents 81,737 26,962
Cash and cash equivalents, beginning of period 82,230 46,897
Cash and cash equivalents, end of period $ 163,967 $ 73,859

INNODATA INC. AND SUBSIDIARIES
RECONCILIATION OF GAAP TO NON-GAAP FINANCIAL MEASURES
(Unaudited)
(In thousands)

Adjusted Gross Profit and Adjusted Gross Margin

For the Three Months Ended September 30, For the Nine Months Ended September 30,
2026 2025 2026 2025
Gross Profit attributable to Innodata Inc. and Subsidiaries $ 35,537 $ 25,504 $ 117,789 $ 71,779
Depreciation and amortization 2,199 1,729 6,559 4,855
Stock-based compensation 597 444 1,943 1,312
Adjusted Gross Profit $ 38,333 $ 27,677 $ 126,291 $ 77,946
Gross Margin 37 % 41 % 42 % 40 %
Adjusted Gross Margin 40 % 44 % 45 % 43 %

Adjusted EBITDA

For the Three Months Ended September 30, For the Nine Months Ended September 30,
2026 2025 2026 2025
Net income attributable to Innodata Inc. and Subsidiaries $ 10,329 $ 8,342 $ 39,639 $ 23,348
Provision for income taxes (340) 3,837 5,247 6,718
Interest income, net (1,334) (420) (3,471) (1,124)
Depreciation and amortization 2,256 1,748 6,730 4,913
Stock-based compensation 6,814 2,707 19,918 8,309
Adjusted EBITDA $ 17,725 $ 16,214 $ 68,063 $ 42,164

SOURCE: Innodata Inc.

View the original press release on ACCESS Newswire

View All News

Innodata to Report Third Quarter 2026 Results Bull 0 Bear 0 Routine

10/15/2026

NEW YORK, NY / ACCESS Newswire / October 15, 2026 / INNODATA INC. (Nasdaq:INOD) today announced that it will report Third Quarter 2026 results after the market closes on Thursday, November 5, 2026. A news release will be available in both the News and Investor Relations sections of the Innodata website, www.innodata.com .

Innodata has scheduled an investor conference call for 5:00 PM Eastern time on that same day.

The call-in numbers for the conference call are:

(+1) 833 296 5777 North America, Toll Free
(+44) 808 196 8935 United Kingdom
(+1) 206 407 3228 International
Participant Access Code 730 768 717
For Replay:
(+1) 833 309 1852 North America-Toll Free
(+44) 800 007 4044 United Kingdom
(+1) 929 828 5978 International
Access Code 730 768 717

Investors are also invited to access a live Webcast of the conference call at the Investor Relations section of Innodata's website at https://investor.innodata.com/events-and-presentations/ . Please note that the Webcast feature will be in listen-only mode.

Call-in replay will be available for seven days following the conference call, and Webcast replay will be available for 30 days following the conference call, at the Investor Relations section of Innodata's website at https://investor.innodata.com/events-and-presentations/ .

About Innodata

Innodata (Nasdaq: INOD) is a global data engineering company. We believe that data and Artificial Intelligence (AI) are inextricably linked. Our mission is to enable the responsible advancement of artificial intelligence by providing the data, evaluation frameworks, and human expertise required to build AI systems that can be trusted at scale. We provide a range of transferable solutions, platforms, and services for Generative AI / AI builders and adopters. In every relationship, we honor our 36+ year legacy delivering the highest quality data and outstanding outcomes for our customers.

Visit www.innodata.com to learn more.

Forward-Looking Statements

This press release may contain certain forward-looking statements within the meaning of Section 21E of the Securities Exchange Act of 1934, as amended, and Section 27A of the Securities Act of 1933, as amended. These forward-looking statements include, without limitation, statements concerning our operations, economic performance, financial condition, developmental program expansion and position in the AI services market. Words such as "project," "forecast," "believe," "expect," "can," "continue," "could," "intend," "may," "should," "will," "anticipate," "indicate," "guide," "predict," "likely," "estimate," "plan," "potential," "possible," "promises," or the negatives thereof, and other similar expressions generally identify forward-looking statements.

These forward-looking statements are based on management's current expectations, assumptions and estimates and are subject to a number of risks and uncertainties, including, without limitation, impacts resulting from ongoing geopolitical conflicts; anticipated and actual use cases and outcomes; investments in large language models; that contracts may be terminated by customers; projected or committed volumes of work may not materialize; pipeline opportunities and customer discussions which may not materialize into work or expected volumes of work; the likelihood of continued development of the AI markets, particularly new and emerging markets, that our services support; the ability and willingness of our customers and prospective customers to execute business plans that give rise to requirements for our services; continuing reliance on project-based work and the primarily at-will nature of such contracts and the ability of these customers to reduce, delay or cancel projects; potential inability to replace projects that are completed, canceled or reduced; revenue concentration among a limited number of customers; our dependency on third-party providers and partners; our ability to achieve revenue and growth targets; difficulty in integrating and deriving synergies from acquisitions, joint ventures and strategic investments; potential undiscovered liabilities of companies and businesses that we may acquire; potential impairment of the carrying value of goodwill and other acquired intangible assets of companies and businesses that we acquire; a continued downturn in or depressed market conditions; changes in external market factors; the potential effects of U.S. global trade and monetary policy, including the interest rate policies of the Federal Reserve; changes in our business or growth strategy; the emergence of new, or growth in existing competitors; various other competitive and technological factors; our use of and reliance on information technology systems, including potential security breaches, cyber-attacks, privacy breaches or data breaches that result in the unauthorized disclosure of consumer, customer, employee or company information, or service interruptions; and other risks and uncertainties indicated from time to time in our filings with the Securities and Exchange Commission ("SEC").

Our actual results could differ materially from the results referred to in any forward-looking statements. Factors that could cause or contribute to such differences include, but are not limited to, the risks discussed in Part I, Item 1A. "Risk Factors," Part II, Item 7. "Management's Discussion and Analysis of Financial Condition and Results of Operations," and other parts of our Annual Report on Form 10-K, filed with the SEC on February 26, 2026, and in our other filings that we may make with the SEC. In light of these risks and uncertainties, there can be no assurance that the results referred to in any forward-looking statements will occur, and you should not place undue reliance on these forward-looking statements. These forward-looking statements speak only as of the date hereof.

We undertake no obligation to update or review any guidance or other forward-looking statements, whether as a result of new information, future developments or otherwise, except as may be required by the U.S. federal securities laws.

Company Contact

Aneesh Pendharkar
[email protected]

(201) 371-8000

SOURCE: Innodata Inc.

View the original press release on ACCESS Newswire

View All News

By Yonatan Brander, Esther Derman, Avigdor Rosensweig

Published on October 6, 2026

Personalization Needs to Be Trained

Frontier models have made remarkable gains in reasoning, instruction following, tool use, full computer-use pipelines, and deep domain knowledge in code, math, and other areas. Each of those gains came the same way: define the capability, build the training data that exercises it at scale, and train the model on it until the behavior holds. 

What about personalization? 

Today’s models are eager to comply, yet genuine personalization is rarely captured by benchmark leaderboards. It often requires post-training because it does not simply emerge from general competence, and neither a system prompt nor retrieved memory can supply it on its own. 

An agent that adapts to the person it works for needs to learn from long-horizon, personalized tasks at scale. Why is that, and how was Innodata’s platform designed to produce the data needed to support it? 

Generic Answers Are Practically Wrong

People expect an agent to remember what matters to them without being told their whole story every time. 

A business owner who asks for a customer quote expects it at their prices, in their voice, and for a day they are available. If the agent gets those details wrong, the result is useless to the person who asked, even though the task was technically “done.” 

Yet today’s evaluation benchmarks often miss this type of failure: 

  • The success criterion is whether the task was completed. But the same request from two different people can have two different right answers. A personalized success criterion should instead ask: “Would this person, with this history, consider the result correct?” Only the reality of a long-horizon exchange with a human user can supply that judgment. 

  • LLM judges can gravitate toward the average of the human interactions represented in their training data, rather than the months of context that determine whether a response fits a particular person. 

  • Outside evaluators who did not participate in the user-agent relationship may not know what the person wanted, which constraints were implicit, or why a plausible answer was still wrong. Years of a person’s messages contain emotion, negotiation, evolving preferences, and context that cannot always be reduced to names, dates, and explicit facts. 

If the training signal cannot distinguish a personalized answer from a generic one, the model receives little signal pushing it toward personalization. It learns to satisfy the instruction for the median user—the user the available data most clearly describes. 

A memory file at inference time can give the model relevant information in its context window, but it may not teach the model what to weigh, when an old constraint should override a new instruction, or how different pieces of personal context should influence an action. Those behaviors need to be represented in the training data itself. 

Data Curation for Personalization

To capture the uniqueness of each user, personalization data should account for three things: 

  • Long “relational” horizon: A long-horizon task usually means thousands of interaction steps—for example, a computer-use agent working through a sandbox for an hour. In our setup, long horizon means relationship. To serve today’s request, the agent has to recall a conversation from months ago, weigh it against how the person’s thinking has changed since, prioritize among competing pieces of context, and translate the result into action. Personal agents fail on this far more often because training data rarely exercises it. Our tasks require both: multi-step work whose correct execution depends on months of accumulated context with one person. 

  • Implicit constraints: Personal requests are complex because they often contain requirements that are never stated directly. A person whose profile says they never schedule before 10 a.m., travel with a toddler, and keep Friday afternoons free has specific constraints when asking, “Book me an appointment before my flight.” Task completion is necessary, but personalization is what earns the reward and provides the relevant training signal. 

  • No synthetic neatness: A naive way to create personalized training data at scale is to generate it—invent personas, generate tasks, and fine-tune. But synthetic pipelines can struggle to reproduce the inconsistency, contradictions, changing priorities, and individual quirks that make real people difficult to model. Human annotators are important here because they introduce the natural complexity that synthetic personas often smooth away. 

A Bit of Information Theory

The difference between generic and personalized data lies in the information distribution they encode. 

Formally, let a persona be defined as a random variable P. It generates rollouts, modeled as a random variable X that typically depends on P. The total variety in the data decomposes into two entropic terms: 

H(X) = H(X|P) + I(X ; P) 

where: 

H(X | P) is the within-persona entropy, measuring task variability within the same persona environment. 

I(X ; P) is the cross-persona entropy, or the mutual information between persona and task. It measures how much a task rollout informs about the person who generated it. 

Synthetic pipelines can struggle to maximize both terms at the same time, whereas real human data naturally exhibits variability across both dimensions. Caricatured personas may inflate the second term—high mutual information between persona and rollout—while reducing the first through low entropy within the persona, making them predictable. 

This gives a cheap and honest test: try to guess which persona produced a held-out rollout. Its accuracy is a lower bound on I(X; P): Fano’s inequality says it cannot exceed mutual information. If the probe is near chance across personas, the pipeline is not producing enough diversity, whatever the rollouts look like. 

Run the same probe on a real population, and the gap tells you how far the synthetic one has drifted toward the “average user.” 

Mutual information has a known limit: it cannot tell a well-drawn character from a stereotype, since both are highly attributable. Nor can any intrinsic metric say whether a persona stays consistent while drifting across fifty sessions, or whether a synthetic population covers the tail rather than the bulk. 

The most reliable anchor for each of those is a real person saying, “That is not me.” This is why our platform keeps the human at the center rather than treating interviews as merely a bootstrapping step for a generator. 

The Agent Personalization Platform

The Agent Personalization Platform is a training and evaluation environment for personal agents, where all three ingredients above land in the same data packet. 

  • The agent starts its first task with a constructed history grounded in the participant’s life and work, as if it had already worked with this person for half a year. The model must learn to prioritize within a memory that is deliberately messy: conflicting, disorganized, with competing priorities and rules that have exceptions. 

  • We create tasks that depend on participants’ goals, histories, and context, and ask the same people to evaluate the results. A participant is interviewed in depth about their life, work, tools, habits, and how they communicate. That information becomes the agent’s memory and seeds the workspace. Task authors then write requests whose correct answer requires implicit constraints. Finally, the participant reviews the agent’s actions and explains why a result fits or fails. The person who built the context judges the output—the evaluator who can distinguish the personalized answer from the generic one. 

  • The platform runs agents in virtual workspaces with real tools, the user’s data, and a memory store. Under the hood is a fleet of virtual machines, each running a real agent harness plugged into one person’s story: inbox, calendar, messaging, files, CRM, invoicing, seeded to be consistent with that person. The workspace is stateful and mutable, so actions carry consequences in the long run. 

  • Checkpoint technology lets a team freeze memory, tool state, and even search context at a moment in time, change conditions, and rerun the task. Because the environment is controlled and the rubrics are explicit, the same task can be rolled out many times and scored consistently. Rubrics encode the participant’s rules as verifiable checkpoints; their feedback becomes preference data. Both hold across reruns. 

Each task is one self-contained, replayable packet: the persona the agent runs as—identity, user it serves, values, voice, constructed memory, apps it operates in, the tools it can call, and the seeded data behind each—and the task it is asked to complete: an under-specified request, tool specs and backing data consistent with the persona, a rubric of verifiable checkpoints, and the participant’s verdict. 

That structure turns subjective preference into something closer to a verifiable reward. 

A single rollout with a thumbs-up is preference-based, while many rollouts of the same task against a rubric the participant wrote is RLVR applied to human preference. The data packet serves both use cases: preference pairs for RLHF and rubric-scored rollouts for RLVR-style training. 

Concretely, every packet yields both: 

  • Evaluation data: trajectory-level scores on tasks that require months of context beyond completion; rubric checkpoints plus the participant’s verdict and reasons; frozen checkpoints so the same moment is comparable across model versions; and memory ablations that show which recalls break when storage changes. 

  • Training data: preference pairs from the people who hold the preferences; rules turned into verifiable rewards across repeated rollouts; a profile of how each person quotes, books, declines, and writes; and environments that grow richer with every session, sandboxed, portable, and re-runnable. 

Who is This For?

Any team shipping an assistant that is supposed to know its user. 

A shopping agent recommending a laptop to a software developer should account for different priorities than one helping a college student working within a budget. 

A life-sciences research assistant should learn how a particular scientist reads a paper. 

A small-business agent should quote jobs in its owner’s voice. 

A writing assistant should respond the way its author would. 

The recipe is the same in each case: 

  • Bring one user group, or let us recruit a sample of it. 

  • Bring the task your agent gets technically right but practically wrong. 

  • We build the workspace and the checkpoints around it, plug your agent in, and find out which tasks left those people unhappy, what the agent should have called, and which memory it should have consulted. 

  • Train on the annotated answers. 

Personalization is the unlock for agent-as-assistant. 

It is not just a prompt, and it is not just a memory file. It is a training problem—and to solve it, you need to make personalization personal. 

Connect with an Expert

Innodata Inc.

Building an AI agent that needs to truly know its users?

Tell us what you’re building, where personalization is falling short, and what your agent needs to understand over time.

Talk to an Expert

Faceted metallic brain shape centered on a green background, with light concentric rings behind it and small floating pieces nearby.

Get Started

Building an AI agent that needs to truly know its users?

Tell us what you’re building, where personalization is falling short, and what your agent needs to understand over time.

URL

This field is for validation purposes and should be left unchanged.

First Name(Required)

Last Name(Required)

Email(Required)

Tell us what you’re building and where your challenges are.(Required)

Share Your Challenge

Follow Us

Facebook-f Instagram X-twitter Linkedin-in

Innodata wordmark with a blue chip-like icon on the left, placed on a white background with black text.

Innodata (Nasdaq: INOD) is a global data engineering company. We believe that data and Artificial Intelligence (AI) are inextricably linked. Our mission is to enable the responsible advancement of artificial intelligence by providing the data, evaluation frameworks, and human expertise required to build AI systems that can be trusted at scale. We provide a range of transferable solutions, platforms, and services for Generative AI / AI builders and adopters. In every relationship, we honor our 36+ year legacy delivering the highest quality data and outstanding outcomes for our customers.

About

Company

Contact

Facebook Instagram X-twitter Linkedin

By clicking “Accept,” you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. To find out more about the cookies we use, see our Privacy Policy .

AcceptCookie Settings

Manage consent

Close

Privacy Overview

When you visit any website, it may store or retrieve information on your browser, mostly in the form of cookies. This information might be about you, your preferences or your device and is mostly used to make the site work as you expect it to. The information does not usually directly identify you, but it can give you a more personalized web experience. Because we respect your right to privacy, you can choose not to allow some types of cookies. Click on the different category headings to find out more and change our default settings. However, blocking some types of cookies may impact your experience of the site and the services we are able to offer.

Necessary

Necessary

Always Enabled

These cookies are necessary for the website to function and cannot be switched off in our systems. They are usually only set in response to actions made by you which amount to a request for services, such as setting your privacy preferences, logging in or filling in forms. You can set your browser to block or alert you about these cookies, but some parts of the site will not then work. These cookies do not store any personally identifiable information.

Functional

functional

These cookies enable the website to provide enhanced functionality and personalisation. They may be set by us or by third party providers whose services we have added to our pages. If you do not allow these cookies then some or all of these services may not function properly.

Analytics

analytics

Analytics cookies are cookies that track how users navigate and interact with a website. The information collected is used to help the website owner improve the website.

Advertisement

advertisement

These cookies may be set through our site by our advertising partners. They may be used by those companies to build a profile of your interests and show you relevant adverts on other sites. They do not store directly personal information but are based on uniquely identifying your browser and internet device. If you do not allow these cookies, you will experience less targeted advertising.

Save & Accept

Prompts

Innodata Inc. (ticker INOD), listed in United States.

Company History

Date Change Detail
May 16, 2012 Renamed INNODATA ISOGEN INC → INNODATA INC
Mar 18, 2004 Renamed INNODATA CORP → INNODATA ISOGEN INC

ETFs Holding Innodata Inc.

ETF Weight in ETF Shares Market value As of
ASCE 1.81% 28,232 $ 2.134M Jun 30, 2026
SSMG 1.15% 585 $ 36.756K Jul 31, 2026
JSML 0.41% 23,659 $ 1.486M Jul 31, 2026
JSMD 0.38% 63,573 $ 3.994M Jul 31, 2026
PSC 0.33% 102,040 $ 7.712M Jun 30, 2026
FSEG 0.19% 145 $ 10.959K Jun 30, 2026
ESSC 0.13% 742 $ 46.62K Jul 31, 2026
VFQY 0.07% 6,030 $ 340.213K Aug 31, 2026
AVSC 0.06% 60,755 $ 1.98M Nov 16, 2026
VFMO 0.04% 12,722 $ 717.775K Aug 31, 2026
AVUQ 0.02% 2,369 $ 77.206K Nov 16, 2026
DFAS 0.02% 166,435 $ 6.04M Nov 12, 2026
AVSU 0.01% 1,989 $ 64.822K Nov 16, 2026
IALT short -0.01% -23,757 $ -774.241K Nov 16, 2026
MSSM 0.01% 1,092 $ 61.611K Aug 31, 2026
DFSU 0.01% 3,708 $ 134.563K Nov 12, 2026
DXUV 0.01% 9,236 $ 335.174K Nov 12, 2026
DFAC 0.00% 89,636 $ 3.253M Nov 12, 2026
DCOR 0.00% 23,479 $ 852.053K Nov 12, 2026
DFAU 0.00% 6,665 $ 241.873K Nov 12, 2026
DFUS 0.00% 5,379 $ 195.204K Nov 12, 2026
EBI 0.00% 2 $ 112.84 Aug 31, 2026
AVUS 0.00% 20,512 $ 668.486K Nov 16, 2026