Our verdict
NineTwoThree AI Studio is an AI development and consulting firm based in Danvers, Massachusetts, north of Boston (Clutch). Andrew Amann and Pavel Kirillov lead it, and the company says it has been building software products since 2012 and has delivered more than 150 projects for 75 clients with a team of more than 70 people (NineTwoThree).
The firm started as a product design and development studio and now sells AI work first: AI strategy, chatbots and voice assistants grounded in a client’s own data, AI agents, predictive models and computer vision. Its published projects include an AI search experience built on 90 years of Consumer Reports data, a repair-order quality check model for Amerit Fleet Solutions, and an AI content platform for Contently. On Clutch it holds a 4.9 rating from 41 reviews, and clients there describe projects mostly in the $200,000 to $999,999 range.
For a mid-sized company that wants a single team to scope, design and ship an AI product, NineTwoThree offers product managers, designers and machine learning engineers under one roof, along with an unusually deep file of verified client reviews for a firm its size.
The right buyer
- Mid-market companies that want a chatbot, AI agent or predictive model shipped by one product team.
- Buyers with budgets in the low six figures, above its $100,000 Clutch minimum.
- Teams that want a deep record of verified reviews: 41 on Clutch, averaging 4.9.
Reasons to look elsewhere
- Its Clutch minimum of $100,000 puts it out of reach for small pilots and quick throwaway prototypes.
- Two Clutch reviewers raised process issues: slower proof-of-concept cycles than they wanted, and lost context when team members were rotated.
- Much of its review history comes from mobile and web app projects from before its move into AI.
Scores
| Criterion | Weight | Score | Basis |
|---|---|---|---|
| Production track record | 20% | 7.6 | Named AI builds for Consumer Reports, Amerit Fleet and Contently, several confirmed in verified Clutch reviews. |
| Technical depth | 15% | 6.8 | Chat, agents, ML and computer vision; no published research or open-source output. |
| Industry depth | 10% | 6.6 | Work spread across media, fleet services, fintech and health; no single deep vertical. |
| Delivery model and handoff | 15% | 7.6 | Proof of concept, then production build, then training or ongoing support, per its published process. |
| Commercial clarity | 10% | 7.2 | Minimum project size and hourly band are listed on Clutch. |
| Platform independence | 10% | 6.8 | No single-vendor tie stated; Clutch focus lists OpenAI models. |
| Stability and scale | 10% | 7.0 | Building products since 2012; 70+ staff. |
| Independent evidence | 10% | 7.8 | 41 Clutch reviews averaging 4.9, most of them verified. |
| Overall | 100% | 7.2 | Weighted average |
Where NineTwoThree AI Studio fits
| Need | Fit | Why |
|---|---|---|
| Chatbot or AI search on company data | Strong | Consumer Reports and PRC Macro projects. |
| Predictive models and AI agents | Strong | Amerit Fleet quality model; agent work listed. |
| Full product design and build | Strong | Grew out of a product studio. |
| Quick, low-budget prototypes | Weak | $100,000 minimum on Clutch; one client asked for faster POCs. |
What clients and leaders say
“Their ability to learn about our business and understand our goal was truly impressive.”
“POCs and mockups could be developed much faster and allow customers to test them within days rather than weeks or months.”
“On at least one occasion, responsibilities were rotated among engineering team members and some knowledge/context was lost.”