Build production AI systems - not demos. You ll design orchestration, retrieval and evaluation pipelines, route work across model tiers for cost and quality, and ship intelligence that survives real enterprise load.

Visakhapatnam Hyderabad Vijayawada 3-6 years Full time On-site OpenAI Gemini Anthropic

What you ll bring
- 3-6 years building production software, with hands-on LLM or ML work
- Strong Python; model APIs (OpenAI, Gemini, Anthropic) and open models
- RAG, embeddings and vector stores (pgvector, Pinecone or similar)
- Agent orchestration, tool-use, and prompt/context engineering
- Building evaluation harnesses that score correctness, latency and cost
- Serving and integration with FastAPI, Docker and a cloud platform

What you ll do
- Ship orchestration, RAG and extraction pipelines into production
- Route requests across model tiers to balance cost, latency and quality




- Build evaluation gates that decide what ships
- Fine-tune and adapt models where it earns its cost
- Instrument every path with telemetry, tracing and guardrails

Speech / vision pipelines (Whisper, YOLO, OCR) Edge inference (ONNX, TensorRT) Kafka / event-driven systems MLOps and model observability

Shortlisted candidates attend a final in-person round at our Visakhapatnam, Hyderabad or Vijayawada office.

Disclaimer: This job description has been sourced from a public domain and may have been modified by Naukri.com to improve clarity for our users. We encourage job seekers to verify all details directly with the employer via their official channels before applying.

📌 AI Engineer (Hyderabad)
🏢 Tilicho Labs
📍 Hyderabad