Innodata Inc. (INOD)

Key figures

The last twelve months reported, with the latest price and balance sheet.

Revenue, last twelve months

$317.2M

Revenue growth

39.0%

Net income, last twelve months

$46.5M

Earnings per share

$1.32

Operating margin

18.2%

Net margin

14.7%

Free cash flow

$183.7M

Return on equity

29.1%

Market value

$1.6B

Price to earnings

36.1x

Dividend yield

0.0%

Cash

$240.3M

Debt

$0

Debt to equity

0.00

Shares outstanding

34.4M

Financial statements

Revenue, profit and cash flow by fiscal year, each year linked to its annual report.

By quarter

What the company said with its results

From the second quarter 2026 results, filed 6 August 2026.

• Revenue rose 58% year over year. Revenue was $92.1 million against $58.4 million a year earlier; six-month revenue was $182.2 million against $116.7 million.
• Adjusted EBITDA grew 92%. Adjusted EBITDA was 27.5% of revenue. Management called the growth in Adjusted EBITDA operating leverage.
• Net income and diluted EPS rose. Net income was $14.4 million, or $0.41 per diluted share, against $7.2 million and $0.20 a year earlier. Six-month net income was $29.3 million.
• New frontier lab customer landed. The company added a new customer in the quarter, described as one of the fastest-scaling frontier labs.
• Customer base described as broadening. Management said the base is broadening in both customers and customer programs.
• Agentic reinforcement learning program wins. A significant new program with the largest customer covers personalization of long-horizon agents and is now scaling. A second program covers computer-use agent environments.
• Adjusted gross margin rose on mix. Adjusted gross margin was 49%, up from 43%. Management attributed the change to mix toward off-the-shelf datasets and high-value pre-training programs.
• GAAP gross margin also improved. GAAP gross margin was 46%, up from 39% a year earlier, and 45% for the six months against 40%.
• Operating costs rose with revenue. Direct operating costs were $49.7 million against $35.4 million, and selling and administrative expenses were $26.6 million against $14.1 million.
• Full-year revenue growth guidance reiterated. Management reiterated full-year 2026 revenue growth guidance and said the figure does not yet include large potential programs.
• Large programs excluded from guidance. Management described large potential programs with new and anticipated customers as likely wins in its judgment, and said they are not in the growth figure.
• Data collection moves to enterprise scale. Egocentric data-collection pilots with leading robotics companies ran successfully, and the practice moved from single pilots to scoping multi-modal programs at enterprise scale.
• Two benchmarks and cyber suite released. The two benchmarks are meant to surface failure modes that standard leaderboards do not show. The first stage of the AI Cyber Training Suite includes twelve datasets and evaluation systems.
• Federal and enterprise focus named. Abuhoff said he will bring research-enabled capabilities to the federal government and the enterprise. Enterprise AI and Federal strategies are built on technology developed for frontier labs.

Revenue by segment

As the company reports it in the segment note of its annual report.

By business, FY2025

By region, FY2025

Company events

The newest 8-K filings of 186 held.

Every filing

The newest filings with the SEC of 441 held.

Insider trades

Form 4 trades by officers, directors and 10% owners, last 12 months.

Planned insider sales

Form 144 notices, filed when the sell order goes to the broker.

Large holders

Schedule 13D and 13G filings by holders of more than 5% of the voting shares.

Short interest

Shares sold short and not yet bought back, on each FINRA settlement date.

Funds that hold it

From each fund's latest Form 13F.

316 funds held it at 30 Jun 2026: 163 added and 90 trimmed in the quarter. The largest:

Peers

Companies in the same industry.

• Automatic Data Processing Inc. (ADP): revenue $21.9B, market value $113.5B
• Workday, Inc. (WDAY): revenue $10.2B, market value $50.3B
• Reddit, Inc. (RDDT): revenue $2.8B, market value $27.2B
• Verisk Analytics, Inc. (VRSK): revenue $3.1B, market value $24.7B
• Toast, Inc. (TOST): revenue $6.8B, market value $18.9B
• RingCentral, Inc. (RNG): revenue $2.6B, market value $7.6B
• Applied Digital Corp. (APLD): revenue –, market value $6.1B
• Pegasystems Inc. (PEGA): revenue $1.7B, market value $5.8B
• Bilibili Inc. (BILI): revenue $4.2B, market value $5.4B
• SharonAI Holdings Inc. (SHAZ): revenue –, market value $4.8B

What the company does

Innodata is a data engineering and AI systems services company. It supports the development, training, post-training, evaluation and deployment of artificial intelligence systems. The company was founded in 1988 and has its principal executive offices in Ridgefield Park, New Jersey. It employs and engages 12,200 professionals across more than 70 countries. It reports three segments: Digital Data Solutions (DDS), Synodex and Agility.

How the business is organised

The company also describes Technology Practices, which turn research into deployable solutions. It says these are internal capability groupings and not separate operating segments.

Research sits in Innodata Labs, which the company describes as its long-term research function with a typical horizon of two to three years. Platform engineering teams build and update the company's platforms.

What it sells

- Training and post-training data development. - Alignment and preference optimization. - Capabilities, alignment and safety evaluation. - AI enablement and operationalization, including support for agentic and tool-using systems.

Evaluation work includes red teaming and adversarial testing, such as probing prompt-injection and jailbreak weaknesses. The company also does cybersecurity work on large language models and AI agents. In 2025 it launched a Federal Practice aimed at United States government agencies and contractors.

Synodex turns medical records into structured digital data for insurance and healthcare workflows. Its revenue is seasonal, lowest in the third quarter and highest in the fourth, because it follows the number of life insurance applications that insurers receive. Agility provides media intelligence and public relations workflow software with AI-driven monitoring, analytics and content tools, and its revenue comes from subscriptions.

Who buys from it

Customers span banking, insurance, financial services, technology, digital retailing, and information and media. The customer base includes five of the so-called "Magnificent Seven" technology companies and several AI research labs and model developers. Customers outside the United States made up approximately 16% of total revenues.

Work is usually set up under master service agreements, with statements of work, service orders or purchase orders covering the scope, pricing and duration of each project. These arrangements generally do not oblige customers to buy in future periods. Customers can generally end agreements on 30 to 90 days' notice.

The company sells mainly through a direct sales model, with sales staff in the United States, Canada, the United Kingdom and Europe. It is also developing partner and channel relationships.

Where it delivers work

The company says its global footprint allows multilingual work and follow-the-sun delivery. Most revenue is in United States dollars, with a portion in Canadian dollars, Pound Sterling and Euros.

Data and applications are hosted in the company's own data centers, in customers' data centers, and on Amazon Web Services, Microsoft Azure, Oracle Cloud Infrastructure and Google Cloud Platform. Its operations facilities in Asia and its New Jersey facility are certified to the ISO/IEC 27001:2022 information security standard. For work involving protected health information, it uses United States-based co-located data centers or HIPAA-compliant cloud infrastructure. It says it seeks to comply with HIPAA and HITECH, the European Union General Data Protection Regulation, and the United Kingdom General Data Protection Regulation.

How it competes

The company names AI data and evaluation providers as competitors, including Appen, CloudFactory, Surge AI, Invisible Technologies, Turing, Mercor, TELUS Digital and Scale AI. It also competes with broader services firms such as Accenture, Cognizant Technology Solutions, Genpact Limited, Infosys Limited, PwC and Tata Consultancy Services. Some customers may do certain work in-house instead.

It says it competes on service quality, technical depth, scalability, information security and the ability to support complex AI programs over time. It protects its methods mainly through trade secrets, confidentiality agreements, licensing, copyright and trademark law. It holds one patent with about two years of remaining term. Customers generally keep ownership of their data, models and customer-specific deliverables, while Innodata keeps its pre-existing methods and platforms.

Its people

As of 31 December 2025, the company had 10,107 employees, 10,020 of them full-time. Many hold advanced degrees in fields such as law, business, technology, medicine and social sciences. No employees are represented by a labor union. Since 2016 the company's i-Hope Program has built 20 computer labs and 66 smart classrooms across India, Sri Lanka and the Philippines.

*From Innodata's annual report on Form 10-K for the year ended 31 December 2025. [The filing](https://www.sec.gov/Archives/edgar/data/903651/000110465926020655/inod-20251231x10k.htm)*

From the company's 10-K.