AI Engineer

We're building the intelligence layer for Indian process industries.

IndustryOS® turns a factory into something legible process, productivity, quality, EHS, and ESG read as one operational reality and then lets it earn its intelligence. You'll build the AI that makes that real, on data that comes from plants, historians, and safety-critical operations, not just documents.

This is a hands-on, senior/founding-level role. You own workflows end-to-end and ship. If you want a narrow implementation lane, this isn't it.

Experience 1-3 years

Location – Cyber City, Gurugram [On Site]

What you'll build

•     LLM and agentic systems that reason over operational data DCS/historian tags, time-series, sensor streams, and ISA-95-structured objects not only PDFs.

•     Retrieval and prediction over messy industrial reality: engineering documents, P&IDs, incident reports, schematics, plus tag-level plant data.

•     AI features where a wrong answer is a safety and liability event, not a UX bug. Correctness, abstention, and human-in-the-loop are the job.

Responsibilities

1.     Design, build, and iterate LLM-powered workflows retrieval, routing, tool use, function calling, multi-step agents.

2.     Implement agentic apps that plan, call tools/APIs, and maintain state across tasks.

3.     Build RAG pipelines end-to-end: ingestion, chunking, embeddings, indexing, and latency-optimized retrieval including layout-aware parsing of drawings and reports.

4.     Own prompt engineering and evaluation: golden datasets built with domain experts, A/B tests, guardrails, and metrics across latency, cost, quality, and safety.

5.     Productionize with observability (traces, tokens, failures), cost controls, and fallbacks (LangSmith / Langfuse / Arize or equivalent).

6.     Ship backend services and APIs (Python / FastAPI) integrating with data stores, vector DBs, and time-series sources.

7.     Handle deployment reality for industrial clients: VPC, on-prem, or air-gapped environments and self-hosted open models where data can't leave the plant.

8.     Collaborate with PM/Design and SMEs to translate requirements into reliable, safe, user-facing features.

Must have

•     Hands-on with LLMs (OpenAI, Claude, Llama, etc.) and orchestration frameworks (LangChain, LlamaIndex, or custom).

•     Strong Python; RESTful services; clean, tested code.

•     RAG in production vector databases (Pinecone, Weaviate, FAISS, Qdrant) and embeddings.

•     Solid grasp of agent patterns: tool calling, planning/execution, memory, workflow engines.

•     Prompt design, safety/guardrails, and evaluation frameworks with a real point of view on hallucination control and confidence/abstention.

•     Cloud and deployment basics (AWS/GCP/Azure), Docker, Git, CI/CD.

•     Strong debugging mindset and a bias to ship.

Strong plus

•     MCP we lean on it heavily; real familiarity is a signal we're actively looking for.

•     Self-hosted / on-prem inference: vLLM, quantized models, private deployment.

•     Time-series, anomaly detection, or classical ML the predictive side of industrial AI, not just LLM apps.

•     Document AI and vision: OCR, layout parsing, multimodal models for drawings and reports.

•     Model adaptation for cost/quality: fine-tuning, LoRA, distillation, small-model routing.

•     Exposure to manufacturing, process safety, or industrial data.

Why this role

You'll build AI for a domain almost nobody has cracked Indian process industries, at the frontier, with real clients and real plant data. The problems are hard because they're safety-critical and the data is raw. That's the point.