Roles We're Actively Hiring
Apply directly below — attach your resume and we'll be in touch.
Loading open positions…
What Makes This Different
“We don't hire people to write training notebooks. We hire people to ship production AI systems that run inside hospitals and banks.”
Capability Over Credentials
We don't care where you studied. Show us what you've shipped.
Ship Production Code
We measure output by what runs in production, not what passes in a notebook. If it can’t handle real load, real data, and real edge cases, it doesn’t ship.
Own the Full Stack
There’s no “that’s not my job” here. You’ll debug CUDA out-of-memory errors, optimize Docker containers, fine-tune models, and fix CSS in the same week.
Data Sovereignty First
Every system we build keeps client data on client premises. This isn’t a talking point — it’s an architectural constraint that shapes every decision we make.
What You'll Work With
Real infrastructure. Real models. Real GPUs sitting in server racks we maintain.
GPU Cluster: 4 nodes — LLM inference (Qwen3-Omni-30B + TTS), training (ASR, VL, embeddings, MedGemma), secondary LLM (Qwen3.5-35B-A3B), and the MCMS application platform. You'll have direct access.
Small Team. Real Ownership.
No layers of management. No sprint ceremonies for the sake of process. Build, ship, maintain.
Direct Client Interaction
You’ll talk to clients directly — not through PMs or BAs. Understand the problem first-hand, then build the system to solve it.
Deploy to Production Weekly
We ship to production environments every week. Not staging, not dev — real client infrastructure. You’ll see your code running in hospitals and banks.
Progressive Autonomy
Same model we use for our AI agents: start with oversight, earn independence through consistent delivery. Senior contributors own entire product lines.
Jamshedpur-Based
We’re headquartered in Jamshedpur, Jharkhand. Lower cost of living, focused work environment, and direct access to the GPU cluster. Remote considered for exceptional candidates.
Send Your Work
No Resumes. Show Us What You've Built.
We don't read CVs — we read code, system architectures, and deployment logs. Send us links to your best work: GitHub repos, deployed projects, blog posts about systems you've built, or a short write-up of a hard technical problem you solved.
Send Your Work
Email links to your projects, repos, or a write-up of your best technical work.
Technical Review
We review your work asynchronously. If it’s a fit, we’ll schedule a technical conversation.
Paid Trial Project
A short, paid project on a real problem from our stack. Mutual evaluation — you assess us too.