Software Engineer - AI Research

Overview

Design and deliver AI systems that turn enterprise process knowledge, documents, interviews, and system logs into governed, evidence-backed process ontologies and continuously learning intelligence layers.

What you'll do

• Build document ingestion pipelines handling layouts, tables, scanned diagrams, and long-form embeddings.
• Design schema-constrained LLM extraction with validation and automated repair loops.
• Develop retrieval systems using hybrid search, reranking, and query expansion.
• Build Neo4j or equivalent property-graph and entity-resolution pipelines for GraphRAG.
• Integrate terminology stores and retrieval-in-the-loop generation for grounded outputs.
• Define evaluation datasets, grading rubrics, and regression gates with product and research partners.
• Build multi-agent orchestration and secure MCP/A2A integrations across services.

What you'll need

• Production-grade AI development experience in Python with strong typing, test coverage, and async pipelines.
• Deep knowledge of LLM prompting, schema-driven extraction, and automated repair loops.
• Expertise in dense embeddings, hybrid search, reranking, and query expansion.
• Experience with Neo4j or equivalent property-graph databases, entity resolution, and GraphRAG.
• Experience with grounding architectures, terminology stores, and retrieval-in-the-loop generation.
• Proficiency with LangGraph, OpenAI Agents SDK, secure MCP, or A2A integration patterns.
• Experience defining evaluations, graders, regression tests, and tracing with tools such as Langfuse.

Details

• Location: Bengaluru, India.