Founding Applied Research Scientist (PhD) — Industrial Intelligence Platform
Location : Mumbai (preferable) | Type : Full-time |
WHO THIS IS FOR
PhD (or final-year PhD) in AI / ML / CS (or related) who wants their work running on real production lines. You care about solid theory, and you care even more about measurable impact on uptime, yield, and energy.
WHY THIS ROLE
- Own the 0→1 research agenda for edge inference, time-series modeling, and control—and see it in production quickly
- Work directly with founders and pilot customers across chemicals, pharma, food, steel, and energy
- Early equity with real upside, plus the autonomy to choose methods and tooling to hit performance targets
WHAT YOU'LL DO
Take ideas from paper to plant : ingest → features → inference → control loop → monitoring / rollbackDesign models for non-stationary, multivariate sensor data; quantify uncertainty and detect driftOptimize for edge and latency (quantization, distillation, pruning, ONNX / TensorRT, streaming inference)Define datasets, metrics, and SLOs; build evaluation harnesses and ablations that stickPartner with controls / process engineers to integrate models with existing systemsWHAT WE'RE LOOKING FOR
PhD in AI / ML / CS (or closely related); strong publications or equivalent applied workDepth in one or more : time-series / state-space models, anomaly detection, predictive maintenance, Bayesian / UQ, RL / MPC for control, physics-informed MLSolid software fundamentals : Python / C++, PyTorch / JAX, containers, CI / CD, observabilityExperience with streaming systems and edge deployment is a plus (ONNX, TensorRT, gRPC)Bonus : industrial protocols (OPC UA, Modbus, MQTT), MLOps, model compression / compilersWHAT YOU'LL GET
equity, competitive base, and performance incentives with a clear formulaReal ownership of a product area (edge inference and applied research roadmap)Day-one production impact with access to pilot environments and decision-makersSkills Required
Mqtt, Jax, Predictive Maintenance, anomaly detection, Modbus, GRPC, Pytorch, MLops, Containers, Python