Agentic Ai Engineering Lead Programmer Analyst Bilvantis Technologies Hyderabad
Heavy focus on agentic LLM workflows and LangChain-style tooling; building AI dev tools and production agent systems rather than traditional ML training.
About the Role
Lead Engineer for designing and delivering agentic LLM applications that retrieve enterprise knowledge, use tools, and complete business workflows reliably. Hands‑on role combining architecture, implementation, debugging and mentorship to produce production‑ready agentic AI systems.
Job Description
Role
Lead Programmer Analyst responsible for designing and delivering agentic LLM applications and agents that retrieve enterprise knowledge, call tools, and complete business workflows reliably. Stay hands‑on across architecture, implementation, debugging and mentoring while owning delivery quality.
Key Responsibilities
- Translate business workflows into agent responsibilities, tool interfaces, success criteria and approval boundaries.
- Build LLM applications using effective prompts, structured outputs and context management; choose models based on quality, latency and cost.
- Develop RAG pipelines covering document ingestion, chunking, embeddings, hybrid retrieval, reranking and source‑grounded answers.
- Implement tool‑using workflows with state, memory, checkpoints, human approvals, retries and safeguards against repeated actions or loops.
- Integrate business APIs and data sources; enforce permissions and protect against prompt injection and sensitive‑data exposure.
- Build evaluation datasets, regression checks and traces; debug failures and improve task completion, response time and operating cost.
- Own implementation quality and delivery while coaching junior engineers through design, coding, testing and production troubleshooting.
Requirements
- 5+ years professional software or data engineering experience.
- Hands‑on delivery of LLM applications and agentic AI workflows.
- Strong Python and SQL skills.
- API development experience with FastAPI or Flask and validation with Pydantic.
- Experience with asynchronous service integration, streaming, token limits, prompt design, schema validation and model selection.
- Hands‑on LangChain experience or equivalent frameworks; ability to design fixed workflows and agent decision flows.
- RAG development experience using LangChain or LlamaIndex and vector stores such as pgvector or Qdrant, including filters and retrieval evaluation.
- Implementing tool calling, persistent state, context/memory management, approval flows and robust error handling.
- Experience with AI evaluation, observability, access controls, secure tool execution, Git, testing, containers and cloud deployment.
Preferred
- Exposure to CrewAI or LangGraph for agent orchestration and multi‑agent workflows.
- Experience integrating MCP and building multimodal applications.
- Familiarity with Langfuse, LangSmith or equivalent tracing/evaluation tools and prompt/version management.
- Experience with Hugging Face Transformers for open‑model integration, caching and model routing.
- Demonstrated ability to groom and guide junior engineers via pairing, design discussions, code reviews and feedback.
- Define tasks and acceptance criteria, remove technical blockers, and provide documentation and reusable agent components.
Python SQL FastAPI Flask Pydantic LangChain LlamaIndex pgvector Qdrant Langfuse LangSmith CrewAI LangGraph Hugging Face Transformers Git Containers LLM APIs RAG MCP
Skills
Architectural design Agent design LLM application development Prompt engineering API development Asynchronous integration Schema validation Debugging Evaluation & observability Security & access control Mentoring Testing Cloud deployment Performance and cost optimization