We're a venture-backed AI startup building context-aware agents for the IT Companies. Our team blends deep expertise in data infrastructure, natural language understanding, and domain modeling to transform how business users interact with their data.

We’re developing a platform that enables users to ask complex, data-driven questions in natural language- and get precise, explainable answers instantly. By combining AI planning, schema-aware querying, and ontology-driven reasoning, we turn disconnected enterprise data into actionable insights without dashboards or SQL.

As our Lead Platform Engineer, you'll take full technical ownership of our AI-driven data reasoning system. You’ll design and ship the components that fetch and normalize database schemas, map them to a semantic domain ontology using LLMs, and construct join-path graphs to enable intelligent querying. You’ll lead the build-out of Haystack-based agents that decompose natural language questions, resolve them through ontology-aware planning, and execute SQL or API calls across multiple sources in parallel. You’ll oversee logging and observability pipelines, secure credential handling, and schema versioning. From data profiling to query orchestration and governance, you’ll own both design-time and run-time layers-laying the foundation for a production-grade semantic BI platform.

• 5+ yrs Python (asyncio, typing, packaging, test-driven dev).
• BSc in Computer Science from a recognized university (THIS IS A MUST HAVE)
• Proven delivery of LLM-powered back-ends (RAG, agents).
• Fluency in orchestration frameworks like LangChain, Haystack , Semantic Kernel.
• Deep SQL + database‐engineering skills: PostgreSQL, query optimisation, constraints, indexing.
• Experience designing ETL/metadata pipelines (schemas, lineage, profiling).
• Cloud & infra: AWS (S3, Lambda / Step Functions, Secrets Mgr), Docker, GitHub Actions.
• Excellent written communication; able to turn huge requirement docs into crisp deliverables.

• Insurance or other regulated-data domain experience.
• Familiarity with sqlglot, NetworkX, FAISS Haystack.
• Data Engineering Skills and track record of DWH designing, building and maintaining

• Stack - Python, Haystack, OpenAI, FAISS, Postgres, AWS, Docker
• Agile - weekly sprint planning on Monday.com, daily F2F/Slack stand-ups.
• Hybrid - 2 days from gome. the rest is from our opened-space offices.
  

• Competitive salary + early-stage equity.

1. Intro call – 30 min cultural & role fit.
2. Technical deep-dive – 90 min architecture discussion & live design exercise (no leetcode).
3. Exec chat – roadmap alignment, offer details.