FY2018 – FY2025 · INOD · CIK 903651 · FY ends Dec 31
Innodata Financials
Annual revenue, the DDS / Synodex / Agility segment split, the ~58%-of-revenue single-customer concentration story inside the AI-training-data segment, year-end headcount, and the 1988-founded → 2023-AI-pivot framing — every chart point sourced to a 10-K filing. EDGAR XBRL pulled live; segment values hand-curated and source-tagged.
DDS revenue (FY2025)
$220.8M
88% of total — the AI-training-data segment.
Top customer % of revenue
~58%
Single DDS customer, FY2025. Up from ~48% the prior year.
Year-end full-time employees
10,020
FY2025. Labor-heavy services business.
Stock price (INOD)
Live current price plus a five-year history. Updates during US market hours (9:30 a.m. – 4:00 p.m. Eastern, Monday – Friday) via TradingView's embed widget. The bars on this chart are end-of-day; intraday prices live in TradingView's iframe and are not pulled into this site's data.
Most recent Innodata SEC filings
The freshness frontier — the latest filing of each material form type, fetched live from EDGAR. The 10-K marked “Anchors this page” is the source of every annual value; the others are surfaced for transparency.
4 material filings have been filed since the anchor 10-K above (2 8-K, 1 DEF 14A, 1 ARS). The annual financial metrics on this page (revenue, segment splits, secondary metrics) are still anchored to the 10-K because the 10-Q is interim quarterly data and 8-Ks are individual material events.
Annual revenue
As reported on the consolidated statement of operations in each year's 10-K. Innodata's fiscal year ends December 31. Bars FY2018 – FY2025 are pulled live from SEC EDGAR XBRL (RevenueFromContractWithCustomerExcludingAssessedTax for FY2024 onward, with Revenues as the older-tag fallback). The FY2025 number is anchored by the AI-training-data inflection inside the DDS segment — see the segment chart below.
Show table
| Fiscal year | Revenue | Source |
|---|---|---|
| FY2018 | $57.4M | 10-K filed 2020-03-16 |
| FY2019 | $55.9M | 10-K filed 2021-03-15 |
| FY2020 | $58.2M | 10-K filed 2022-03-24 |
| FY2021 | $69.8M | 10-K filed 2023-02-24 |
| FY2022 | $79.0M | 10-K filed 2024-03-04 |
| FY2023 | $86.8M | 10-K filed 2025-02-24 |
| FY2024 | $170.5M | 10-K filed 2026-02-26 |
| FY2025 | $251.7M | 10-K filed 2026-02-26 |
Revenue by segment — DDS, Synodex, Agility
Innodata reports three segments. Digital Data Solutions (DDS) — Digital Data Solutions (DDS) — the AI-training-data segment. Sells annotation, evaluation, alignment, and post-training data services to the foundation-model labs and Big Tech companies that build large language models, plus AI-enabled data engineering for enterprise customers. The single load-bearing customer (~58% of FY2025 total revenues) sits inside DDS. Synodex — Synodex — a medical-records-to-structured-data platform serving life-insurance underwriters and healthcare customers. A small, stable, non-AI segment compared to DDS. Agility — Agility — a public-relations and media-intelligence SaaS platform. Independent of the AI training-data narrative; competes with the broader PR-tech category. The split is hand-curated from each year's 10-K Segment Information note; the EDGAR companyfacts JSON does not expose the disaggregated breakdown. Source link on each row points to the underlying 10-K.
Show table
| FY | DDS | Synodex | Agility | Source |
|---|---|---|---|---|
| FY2021 | $52.6M | $4.2M | $13.0M | 10-K filed 2023-03-01 |
| FY2022 | $56.5M | $7.1M | $15.4M | 10-K filed 2024-02-29 |
| FY2023 | $61.6M | $7.5M | $17.7M | 10-K filed 2024-02-29 |
| FY2024 | $141.1M | $7.9M | $21.5M | 10-K filed 2026-02-26 |
| FY2025 | $220.8M | $7.3M | $23.5M | 10-K filed 2026-02-26 |
Segment income (loss) before income taxes
Innodata's segment-profitability measure under ASC 280 is “Income (loss) before provision for income taxes” after elimination of inter-segment profits — the page treats this as the segment operating-income equivalent. The grouped-bar layout with a heavy zero baseline makes loss-years explicit. The DDS turn from near-breakeven in FY2021 – FY2023 to nine-figure profitability in FY2025 is the AI-customer story; Agility's loss profile through FY2022 (PR/media-intelligence platform investment) flips to a small profit in FY2024 and a near-breakeven result in FY2025.
Show table
| FY | DDS | Synodex | Agility | Source |
|---|---|---|---|---|
| FY2021 | $5.9M | -$1.5M | -$5.4M | 10-K filed 2023-03-01 |
| FY2022 | $716K | -$2.6M | -$8.6M | 10-K filed 2024-02-29 |
| FY2023 | $1.3M | $219K | -$1.3M | 10-K filed 2024-02-29 |
| FY2024 | $21.2M | $1.9M | $1.2M | 10-K filed 2026-02-26 |
| FY2025 | $39.4M | $626K | -$155K | 10-K filed 2026-02-26 |
Customer concentration — the headline story
From the “Concentration of Risk” / “Customer concentration” disclosure of the most recent 10-K. This is the single most important fundamentals fact on the page: a single customer inside the DDS segment supplies the majority of total revenue, and the share has risen over the last two fiscal years.
For the fiscal year ended December 31, 2025, one customer in our Digital Data Solutions (DDS) segment accounted for approximately 58% of the Company's total revenues. For the fiscal year ended December 31, 2024, one customer in the DDS segment accounted for approximately 48% of the Company's total revenues. No other customer accounted for 10% or more of total revenues in either of these periods.
Source: 10-K covering FY ending 2025-12-31, filed 2026-02-26. Innodata does not name the customer; this page does not name them either.
Big-Tech / Magnificent-Seven framing
Innodata says its customer roster includes 'leading AI innovation labs and Big Tech companies (including five of the so-called Magnificent Seven).' The framing has appeared in both 10-K narrative and earnings-call transcripts. Innodata does not name the specific Big Tech customers; the page therefore does not name them either.
Source: 10-K filed 2026-02-26.
Revenue by geography
Revenues from customers located outside the United States represented approximately 16% of total revenues for FY2025 and approximately 21% for FY2024. The US share rose as the AI-training-data customer expanded.
| Region | FY2025 | FY2024 |
|---|---|---|
| United States | 84% | 79% |
| International | 16% | 21% |
Source: 10-K filed 2026-02-26. Country of contracting, not country of end-use.
Other metrics
Operating income, net income, stock-based compensation, and operating cash flow — the standard 10-K-derived secondary metrics, all live from SEC EDGAR XBRL. The FY2024 / FY2025 inflection is the AI-customer scaling visible across every line. Innodata does not file ResearchAndDevelopmentExpense as a standalone XBRL tag — R&D-equivalent activity (Innodata Labs / platform engineering) is described narratively in the 10-K and partly capitalized into platform development costs, so a separate R&D chart is intentionally omitted.
| Metric | FY2019 | FY2020 | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|---|---|---|
| Operating Income (Loss) | — | — | — | — | $318K | $24.3M | $39.9M |
| Net Income (Loss) | $617K | $617K | -$1.7M | -$11.9M | -$908K | $28.7M | $32.2M |
| Stock-Based Compensation | $836K | $913K | $1.8M | $3.3M | $4.0M | $4.0M | $11.1M |
| Operating Cash Flow | $4.3M | $5.7M | $5.2M | -$1.2M | $5.9M | $34.9M | $46.8M |
Founded 1988 · pivoted 2018–2023
Innodata was founded in 1988 as a traditional content-and-data services company — its early business was data conversion and content engineering for publishers, financial-services firms, and information-services companies. The pivot into AI training data started in earnest in 2018 with the introduction of platforms targeting machine-learning data preparation, and accelerated sharply in 2023 when foundation-model labs began contracting Innodata for large-scale annotation, evaluation, alignment, and post-training data work. Revenue roughly doubled from FY2023 to FY2024 and again grew about 48% from FY2024 to FY2025 on the same underlying AI-segment momentum. The 1988 founding and the 2023+ AI-driven inflection are why this page exists as financials rather than a company-history page — the pivot is the story.
Methodology & data sources
Live data. Annual revenue, operating income, net income, stock-based compensation, and operating cash flow are sourced live from Innodata's SEC EDGAR XBRL company-facts endpoint. If EDGAR is unreachable when the page is being prepared, the page is not updated, so visitors never see stale fallback numbers.
Hand-curated, source-tagged. The DDS / Synodex / Agility segment revenue and segment income (loss) split, the customer-concentration paragraph, the geographic revenue split, year-end headcount, and the AI-pivot framing are not exposed in the EDGAR companyfacts JSON. Each value is hand-curated from the underlying 10-K and tagged with the filing's accession number. When a newer 10-K is filed than the one a value was last verified against, a “Verification due” banner appears next to the affected section.
Segment-profitability measure. Innodata's segment-profit metric under ASC 280 is “Income (loss) before provision for income taxes” after elimination of inter-segment profits, not the more conventional segment-operating-income measure. The page treats this as the segment OpInc equivalent because the difference is immaterial at Innodata's scale (minimal non-operating items at segment level).
No R&D chart. Innodata does not file ResearchAndDevelopmentExpense as a standalone XBRL tag. R&D-equivalent activity (Innodata Labs / platform engineering) is described narratively in the 10-K and partly capitalized into platform development costs, so a separate R&D chart on this page would either misrepresent zero or require synthesizing a value from notes. The page omits the chart intentionally.
Customer naming discipline. Innodata describes the 58%-of-revenue customer only as “one customer in our DDS segment” and refers to its broader Big-Tech roster as “leading AI innovation labs and Big Tech companies (including five of the so-called Magnificent Seven).” This page does not name individual customers Innodata has not named in a public document.
Refresh cadence. Hand-curated segment, customer-concentration, geographic, and headcount values are re-verified quarterly, aligned with Innodata's 10-Q filing cadence (early May / August / November following each quarterly period, plus the February 10-K covering the December fiscal-year end).
Fiscal-year handling. Innodata's fiscal year ends December 31. A reference to “FY2025” on this page means the fiscal year ended 2025-12-31. Every chart label uses this convention.
What's intentionally not here. Live market cap (different licensing for market-data fields beyond price; the page refreshes quarterly, so a real-time market-cap calc against fresh share counts would be misleading). Forward guidance, analyst estimates, and price targets (out of scope — the page is reported actuals). Editorial framing about whether the AI-segment growth is sustainable, or whether the customer-concentration risk is acute. Comparison to private-market AI-training-data peers (Scale AI, Surge AI, etc.) — private-vs-public is apples-to-pears. Detailed corporate history beyond the 1988-founding and 2023-pivot framing. The stock-price chart at the top of the page is the only live-market data on the page; it is loaded directly into your browser from TradingView's CDN, not redistributed by this site.
Cross-references. Tech Filings for every Innodata SEC filing in chronological order. AI-pure-play organizations for the focused subset of public US-listed AI companies that includes Innodata. Innodata's row on the Orgs index for the company profile.
Last updated: 2026-04-30 18:20 UTC. Latest 10-K on EDGAR: 0001104659-26-020655 filed 2026-02-26. Hand-curated source-tag anchor: 0001104659-26-020655.