Our verdict
ThirdEye Data has built data and AI systems since 2011, has delivered more than 50 solutions and accelerators as a direct Microsoft vendor, and now sells more than 20 pre-built AI solutions that it customizes on a client’s data. It suits a manufacturer, utility, warehouse operator, or IT team that wants a proven vision, maintenance, or document AI system adapted and running in four to six months at offshore rates, and it is less suited to a buyer who wants an all-US team or a strategy-led transformation program.
The right buyer
- Manufacturers, utilities, and warehouses that want vision or maintenance AI on existing cameras and sensors. Vision Count runs on current CCTV and predictive maintenance reads existing sensors (ThirdEye Data).
- Microsoft shops building on Azure. It has worked with Microsoft for more than 15 years across Azure OpenAI, AI Foundry, Fabric, and Copilot Studio (ThirdEye Data).
- Buyers who want to test before committing. It runs a demo hub of more than 20 solutions and offers proofs of value on historical data (ThirdEye Data).
Reasons to look elsewhere
- Most engineering runs from India, so a buyer who needs an onshore team should ask about staffing.
- Its model starts from its own pre-built solutions, so a buyer wanting a fully bespoke research build should compare custom AI shops.
ThirdEye Data, At A High Level
Dj Das founded ThirdEye in Silicon Valley as Third Eye Consulting Services to deliver big data consulting, after business development and technical roles at LG CNS America, Siebel Systems, and Microsoft (ThirdEye Data). Aparajeeta Das, a business intelligence veteran, is co-founder, and Phil McCluskey is COO (ThirdEye Data). The firm moved from big data engineering into data science and then AI, and now positions itself as a full-stack AI development company with head office in San Jose and delivery centers in Kolkata, Hubli, and Brossard (ThirdEye Data).
- 105 employees, up 54% in a year, according to Growjo; we estimate annual revenue at about $8 million (Growjo).
- More than 20 pre-built solutions across vision, generative AI, agents, and forecasting, with typical go-live in 120 to 180 days (ThirdEye Data).
- ISO 27001 and SOC 2 Type 1 certified, with on-premise or private cloud deployment (ThirdEye Data).
- Published work includes aircraft component failure prediction, anomaly detection on electric poles, and a multi-agent loyalty system for a marketing agency (ThirdEye Data).
Scores by criterion
We scored ThirdEye Data from its homepage, leadership page, founder interviews, and Growjo workforce data. Technical depth and production record score highest; independent evidence scores lower because its public record is mostly its own case pages and client testimonials.
| Criterion | Weight | Score | Basis |
|---|---|---|---|
| Production track record | 20% | 6.4 | Production systems since 2011 and more than 50 Microsoft deliverables. |
| Technical depth | 15% | 7.0 | Computer vision, predictive maintenance, document AI, and multi-agent work. |
| Industry depth | 10% | 6.5 | Manufacturing, utilities, logistics, telecom, and financial services. |
| Delivery model and handoff | 15% | 6.5 | Dedicated project managers; offshore delivery from India. |
| Commercial clarity | 10% | 6.6 | Pre-built solutions cut setup cost; 120 to 180 day timelines. |
| Platform independence | 10% | 6.6 | Builds on Azure most often but deploys on-premise and in private clouds. |
| Stability and scale | 10% | 6.5 | Founded 2010; about 105 staff. |
| Independent evidence | 10% | 6.0 | Attributed client testimonials from Southern California Edison, Glas Trösch, and Data.World. |
| Overall | 100% | 6.5 | Weighted average |
Where ThirdEye Data fits
ThirdEye Data is used by mid-sized and large companies in manufacturing, utilities, logistics, telecom, and IT that want a working AI system, such as stock counting from CCTV, failure prediction from sensors, or document extraction, without paying for a long research build. On price it sits in the offshore-weighted lower tier, below US onshore AI shops. Buyers hiring an offshore-priced team for this kind of applied AI build also compare Fusemachines for forecasting and predictive maintenance, Intuz for AI agents and automation from India, and InData Labs for computer vision and predictive models.
| Need | Fit | Why |
|---|---|---|
| Computer vision on existing cameras | Strong | Vision Count, PPE detection, and quality inspection run on current CCTV. |
| Predictive maintenance from existing sensors | Strong | Aircraft and utility asset cases. |
| Document extraction and AI search | Good | Optira document automation and knowledge systems. |
| An all-onshore US team | Weak | Delivery centers are in India and Quebec. |
What clients and leaders say
ThirdEye Data’s leaders and clients describe its work in published announcements. One buyer concern is written from public facts about the firm.
“ThirdEye Data was adaptive to our changing requirements and successfully met model performance metric targets. The team delivered most of the ML models on time, and they were very collaborative and supportive throughout the engagement.”
“ThirdEye Data planned the project well, completed the work on time, and delivered what we expected. It was important that a partner matched the deliverables, timeline, and budget, and ThirdEye Data met all those requirements, which is rare in most projects.”
“We had to make sure that their level of engineering quality met a certain level; they couldn’t just throw warm bodies in our seats.”
“The models hit their targets, but a couple of milestones slipped while the team worked through problems in newer tools, and we had to plan around that.”
Where customer reviews couldn’t be found, our analysts attempted to give a voice to buyer sentiment, drawing on real information from review sites, forums and product documentation. Based on: Dj Das’s Gigstartup interview, in which he lists timeliness on newer technologies as client feedback.
Who leads ThirdEye Data
Dj Das founded ThirdEye Data and is its CEO; he previously worked at Microsoft, Siebel Systems, and LG CNS America (ThirdEye Data).

