Job Description
Key Responsibilities
- Design and develop LLM-powered RAG systems using enterprise documents to build searchable knowledge bases.
- Build AI agents capable of analyzing insurance/reinsurance submissions and identifying required next actions, missing information, and relevant documents.
- Develop agentic workflows using LLM tool/function calling and frameworks such as LangGraph, LangChain, LlamaIndex, Semantic Kernel, etc.
- Build tools for agents to interact with vector databases, SQL/enterprise databases, APIs, web search, and internal applications.
- Develop multi-agent systems where different AI agents can communicate, delegate tasks, and collaborate.
- Implement document ingestion, chunking, embeddings, vector search, hybrid search, and reranking.
- Integrate and work with commercial/open-source LLMs and embedding models.
- Develop natural-language database querying / Text-to-SQL capabilities with appropriate security and controls.
- Build feedback and evaluation mechanisms for continuous improvement/self-learning of the AI system.
- Implement monitoring, logging, evaluation, hallucination/grounding checks, and AI security/guardrails.
- Deploy and integrate AI services with existing enterprise applications and databases.
Required Skills
- Strong Python and backend development experience.
- Hands-on experience with LLMs, RAG, vector databases, and AI agents.
- Experience with LangGraph/LangChain/LlamaIndex or similar frameworks.
- Experience with tool calling, function calling, agent orchestration, and multi-agent systems.
- Strong SQL/database and API integration experience.
- Experience with PostgreSQL/SQL Server/Oracle and vector databases.
- Experience with cloud platforms such as Azure/AWS/GCP.
- Experience building production-grade GenAI applications, not just prototypes.
- Knowledge of insurance/reinsurance, underwriting, submissions, T&A, or related workflows is a strong plus.
Location: DHA Phase 3 Lahore
Timings: 1pm - 10 PM
