What we build: Physical AI breaks after deployment, models drift, conditions change, and on the edge there's no cloud to catch it. Edgeble's self-correcting runtime keeps deployed AI accurate on-device, without stopping inference. It's in production at Tier-1 manufacturers, and hardware-agnostic across edge NPUs.

The role: own the model layer of the runtime. How a deployed model's degradation is detected, how it's adapted on-device under tight compute and memory budgets, and how you prove an adapted model is better before it ever serves. This is the hard, unsolved part of edge AI: adaptation that's protected enough for production lines. You'll take a working system further toward more model families, tighter resource envelopes, and field-grade robustness.

How we work, read before applying: Edgeble is agent-native. We build with agentic coding workflows daily on internal platform tooling already set up for it,



you direct the agents; your judgment goes on what agents can't do: correction logic, validation design, what "better" means. If you'd rather type every line yourself, this role will frustrate you self-select accordingly.

You: 5–6 years of genuine ML engineering depth, training and fine-tuning (not only inference integration), quantization for edge targets, comfort with resource-constrained deployment. Embedded exposure a plus. Concrete evidence you work well agent-augmented we'll ask how, specifically.

Why here: own a defined layer of a production, patent-pending runtime at the moment it scales early-team equity, direct work with a 19 years experience founder from Silicon, Embedded, Edge AI, Linux-kernel/U-Boot maintainers, Physical Systems in the working style most teams are still debating.

📌 Mts On Device Model Correction Hyderabad
🏢 Edgeble
📍 Hyderabad