Applied AI research & consulting lab

AI systems that
survive production.

Most work in the demo and break under real traffic. I design the architecture, evaluations, and infrastructure that hold — then hand your team the keys.

Shipping LLM systems since GPT-3 · 300+ engineers trained

Booking Q3 2026 · limited engagements
trace · answer.rag.eval872 ms
RetrievalGenerationJudgeGate
llm.generate
model
claude-sonnet-4.5
tokens_in
3,140
tokens_out
412
cost
$0.011
gate: PASS

Shipped with evidence — not vibes.

Example trace — a production RAG eval, anonymized. Hover a span.

Trusted by teams who ship

  • Dell Technologies logo
  • Accel logo
  • Razorpay logo
  • HFCL logo
  • Calsoft logo
  • Arize AI logo
  • Ragas logo
  • HoneyHive AI logo
  • Literal AI logo
  • Athina AI logo

From Fortune 500 to Series A — and the AI-infra companies themselves. The common thread: they needed it to work, not just demo.

The record

Not theory. Not decks. Systems that ship.

2020
Since GPT-3

Building production LLM systems before “AI engineer” was a job title.

300+
Engineers trained

The capability stays after the engagement ends. You keep the muscle.

10+
Named clients

Enterprise, fintech, funds, and the eval companies other labs rely on.

From the lab · private beta 2026

evalOS

The eval layer I kept rebuilding for clients, now a product. One line of code turns on OpenTelemetry-native observability, a swapped-position judge panel, and human-in-the-loop quality gates.

Explore evalOS
  • OpenTelemetry-native
    Ingests your existing spans
  • Judge panel
    3 model families, positions swapped
  • Quality gates
    Ship only past your thresholds
  • Cost + drift
    Dashboards that flag regressions

What I do

Three things. Done well.

Production architecture

RAG, agents, and eval loops that hold under real traffic — not benchmark traffic.

From $20K · fixed scope

Technical due diligence

For funds and acquirers: what's actually under the hood, and what breaks at scale.

From $8K · per target

Fractional AI leadership

Senior technical direction through high-growth phases, without the $500K hire.

From $25K / mo

Nothing below $10K. Priced on outcomes, not hours.

How we engage

Two ways in.

Secondment

Embed & execute

I integrate with your team — standups, PRs, Slack — and stay until it ships. Not a deck and a disappearance.

Duration
3 months minimum
Model
Embedded, weekly
Best for
Core builds, overhauls
Contracting

Target & resolve

A specific problem, scoped and solved. Fixed scope, fixed timeline, fixed price. You get the deliverable, not a report about it.

Duration
2 weeks minimum
Model
Fixed scope
Best for
Evals, pre-raise cleanup

The bench

When it needs more firepower.

Not contractors — collaborators. Each could run their own shop.

Rajaswa Patil
Applied AI · ex-Postman, Microsoft Research
Anshul Bhide
AI Strategy · ex-Replit, VC · MIT Sloan
Ayush Chaurasia
ML/CV · co-founded Ultralytics (YOLOv8)
Kumar Shivendu
ML Platforms · Qdrant, search, devtools
Vipul Maheshwari
Applied ML · LanceDB, Superlinked
Rohan Balkondekar
GenAI Lead, Air Arabia · F500 + YC

Ship with evidence.

I take on limited engagements. If it's not a fit, I'll tell you. If it is, we move fast.

Response within 48 hours · no sales process · direct conversation

Transfrm Labs
by Rachitt Shah

Applied AI systems, production-grade. Building with teams at Accel, Sequoia, and friends. Bangalore · San Francisco.

measured on your device just now →CLS0.000

We hold your systems to the same standard.

© 2026 Transfrm LabsAll systems operational