Business Overview / Sources of Revenue

**Innodata (INOD)** is a global data engineering company specializing in AI data preparation, annotation, model training, deployment, and integration services across segments like Digital Data Solutions (DDS), Synodex, and Agility.[1][2]

It earns revenue primarily by providing AI datasets, human-in-the-loop annotation, generative AI training data, safety testing, and enterprise platforms for industries including tech, healthcare, finance, and defense, serving Big Tech and governments.[1][2][3]

No specific percentage breakdown of revenue sources by segment is available in recent data; total TTM revenue was US$238.47m.[1]

Revenue Growth Potential and Recurrence

Innodata’s revenue base is **increasingly recurring but still meaningfully project‑based**. Management highlights “annual recurring revenue” from multi‑year trust‑and‑safety and model‑evaluation contracts with Big Tech customers, but also warns that most DDS contracts are at‑will and customers can reduce or cancel projects, underscoring limited contractual visibility.[2][3]

Growth has been exceptional: revenue grew from **$86.8M in 2023 to $170.5M in 2024** and then another 120% year‑over‑year in Q1 2025, with 79% and 20% organic growth in Q2 and Q3 2025 respectively.[1][2][4][6][7] Management guided to **≥40% revenue growth for full‑year 2025**.[2][3]

Over the next 5+ years, if Innodata maintains strong Big Tech relationships and AI data demand persists, a **high‑teens to 30%+ annual growth range** appears plausible but dependent on continued AI investment cycles and customer concentration risk (inference based on recent run‑rate and guidance).[1][2][7]

Economic Moat Factors

Innodata (INOD) has a **narrow economic moat** driven by **niche expertise** in high-precision AI data engineering for Big Tech and government contracts, creating **switching costs** for clients reliant on its "smart data" platforms and model-safety frameworks.[2][4] Proprietary tools and partnerships with five "Magnificent Seven" firms, plus $68M pre-training deals, hint at **unique assets** and potential **network effects** in the AI supply chain.[2][4]

However, **client concentration** (61% from one customer) undermines durability, with limited **brand power**, **economies of scale**, or insurmountable barriers vs. rivals like Cognizant.[2] High growth (45%+ 2025 revenue guidance) relies on volatile AI demand, not a wide moat.[4] (98 words)

Leadership

**Jack Abuhoff** serves as Innodata's (INOD) **CEO and Chairman**. He is not a founder (company founded 1988), and tenure details are unavailable from sources. Ownership stake is not disclosed. Recent 2025 changes include **Rahul Singhal** as President & Chief Revenue Officer, **Daniel H. (Don) Callahan** and **Gen. (Ret.) Richard D. Clarke** as new independent directors, and **Stewart Massey** as Lead Independent Director. Other execs: Amy Agress (General Counsel), AK Mishra (EVP & COO).[1][3][4][5][6] (78 words)

Financial Health

Innodata’s balance sheet is **very healthy**: it is **debt-free** with about **$73.9M in cash and short-term investments**, so cash-to-debt is effectively infinite.[2]

The company is now a **consistent free cash flow generator**, with strong working capital and cash build, although precise recent **FCF margin** is not explicitly disclosed in sources provided.[1][2]

Innodata has been **net dilutive** over time, issuing shares rather than meaningfully repurchasing; there is no evidence of a sustained buyback program.[2]

Last updated Dec 15, 2025