Job Title: Principal AI Architect Generative AI & Agentic Systems
Team: Analytics & AI / Workforce Solutions
We are looking for a hands-on technical leader to define the future of enterprise AI at EXL. We are moving beyond standard chatbot implementations to build autonomous agentic systems and complex reasoning engines.
In this role, you will not just architect solutions; you will lead the design of next-generation frameworks that combine Graph RAG, Temporal Knowledge Graphs (TKG), and Multi-Agent Orchestration to solve high-value workforce and operational challenges. You will act as the technical bridge between client executives and our deep-tech engineering teams, ensuring we deliver SOTA (State-of-the-Art) solutions that scale.
Key Responsibilities
1. Advanced GenAI Architecture & Innovation
- Architect Agentic Systems: Design and deploy multi-agent workflows where AI agents can plan, reason, and execute tasks autonomously (using frameworks like LangGraph, AutoGen, or custom orchestrators).
- Next-Gen RAG Implementation: Move beyond naive vector search. Implement Graph RAG and hybrid retrieval strategies that combine semantic search with structured knowledge to reduce hallucinations and improve context awareness.
- Knowledge Graph Integration: Lead the development of Temporal Knowledge Graphs (TKG) to capture dynamic relationships and time-series context, enabling AI models to reason across evolving datasets.
- Production-Grade Engineering: Ensure all GenAI solutions are architected for latency, cost-efficiency, and reliability, incorporating semantic caching, guardrails, and robust evaluation frameworks (RAGAS, TruLens).
2. Strategic Technical Leadership
- Value Engineering: Translate ambiguous client workforce challenges into concrete technical roadmaps. Demonstrate how Agentic AI can autonomously handle complex auditing, claims processing, or workforce planning tasks.
- Solution Defense: Act as the primary technical authority in client pursuits. defend architectural choicesexplaining why a Knowledge Graph approach beats standard RAG for specific use casesto CTOs and Chief Data Officers.
- Prototype to Production: Lead the hands-on development of MVPs that prove the value of advanced techniques (e.g., Chat with your Data using TKG) and oversee their transition to scalable production environments.
3. Collaboration & Ecosystem Growth
- Tech Stack Modernization: Drive the adoption of modern AI infrastructure (Vector Databases like Pinecone/Weaviate, Graph DBs like Neo4j, and LLM Ops tools).
- Mentorship: Elevate the technical bar of the wider team. Mentor Data Scientists and ML Engineers on the nuances of prompt engineering, fine-tuning vs. RAG context injection, and agent tool-use.
- Thought Leadership: Represent EXL in the GenAI community, publishing insights on applying Graph RAG and Agentic frameworks to enterprise problems.
What We Are Looking For (The Must Haves)
- Engineering DNA: You have a background in software engineering or ML engineering (ex-AWS, Google, or high-growth tech preferred). You are comfortable reading code and debating implementation details.
- Advanced RAG Mastery: Deep understanding of the limitations of standard RAG. Experience implementing Graph RAG (knowledge graph-enhanced retrieval) and optimizing retrieval pipelines for precision/recall.
- Agentic AI Experience: Proven experience building or prototyping with agentic frameworks (e.g., LangChain Agents, AutoGen, CrewAI). You understand the complexities of agent memory, planning, and tool calling.
- Graph & Data Proficiency: Familiarity with Knowledge Graph construction, ontologies, and specifically Temporal Knowledge Graphs (handling time-sensitive data).
- Business Acumen: The ability to map these complex technologies to measurable business outcomes like 30% reduction in manual audit time or 20% improvement in workforce allocation.
