Principal Architect – Agentic AI

Role: Principal Architect – Agentic & Generative AI (LLM / GenAI Architect)

Experience

                   

Location

Pune / Gurgaon (Gurugram) / Bangalore (Hybrid)

About EXL

EXL is a global leader in data, analytics, and AI-led digital transformation , partnering with global enterprises across Insurance, Healthcare, Banking & Capital Markets, Retail, Media, and Energy. EXL combines deep industry context with advanced AI engineering to deliver measurable business outcomes at scale.

Role Summary

We are seeking a Principal Architect – Agentic & Generative AI to lead the architecture, design, and enterprise-scale adoption of LLM-powered and agentic AI systems .

This role is hands-on and strategic —you will architect complex GenAI solutions, set technical blueprints and standards, mentor senior engineers, and work directly with business and client stakeholders to take GenAI initiatives from POCs to production-grade platforms .

Mandatory: Strong prior experience in Data Engineering or Data Science , enabling deep understanding of data pipelines, model lifecycle, governance, and analytics-driven systems.

Key Responsibilities

GenAI & Agentic AI Architecture

• Define enterprise reference architectures for Agentic AI and LLM-powered platforms , including:

  • Single-agent and multi-agent systems

  • Tool-calling and function orchestration

  • Memory, planning, and execution layers

• Own architectural decisions for Claude / Claude Code and other enterprise-grade LLMs , including model selection, deployment patterns, and cost–latency trade-offs.

• Design secure-by-default GenAI systems incorporating:

  • Guardrails and policy enforcement

  • Data privacy, PII handling, and prompt safety

  • Controlled tool execution in regulated environments

RAG, Knowledge & Data Systems

• Architect large-scale RAG solutions , covering:

  • Data ingestion and curation pipelines

  • Chunking and embedding strategies

  • Vector databases and hybrid search

  • Evaluation and feedback loops

• Partner with Data Engineering teams to ensure data quality, lineage, observability, and governance for AI-driven systems.

Platform & Engineering Excellence

• Drive production readiness of GenAI systems:

  • API-first design (FastAPI / REST / event-driven)

  • CI/CD for LLM workflows

  • Monitoring, evaluation, and cost tracking

• Establish engineering standards, reusable frameworks, and accelerators for faster adoption across EXL accounts.

• Review and influence cloud architecture (Azure / AWS / GCP) for scalable and compliant AI deployments.

Leadership & Stakeholder Engagement

• Act as a technical authority for GenAI across delivery teams and client engagements.

• Mentor senior engineers, tech leads, and architects on agentic patterns and advanced LLM engineering.

• Partner with clients, product owners, and domain SMEs to shape AI roadmaps, solution designs, and value articulation .

Mandatory Skills & Experience

12+ years of total experience with deep hands-on expertise in Generative AI / LLM-based systems , and strong prior background in Data Engineering or Data Science (mandatory) .

Generative AI / LLM Expertise

• Deep hands-on experience with:

  • Claude / Anthropic ecosystem (including Claude Code exposure is a strong plus)

  • Other enterprise LLMs (OpenAI, Mistral, LLaMA, etc.)

• Strong command over:

  • Prompt engineering, prompt orchestration, and agent workflows

  • Tool/function calling, planning–execution loops

  • LLM and RAG evaluation techniques (precision, grounding, faithfulness)

Agentic & RAG Architecture

• Proven experience designing:

  • Agentic AI systems (ReAct, Plan-and-Execute, multi-agent setups)

  • RAG architectures using vector databases (FAISS, Pinecone, Chroma, etc.)

• Strong understanding of hallucination mitigation, guardrails, and safety frameworks .

Core Engineering & Platform Skills

• Expert-level Python engineering (production-grade systems).

• Strong experience with cloud-native AI solutions on Azure, AWS, or GCP.

• API design, microservices, and event-driven architectures.

Mandatory Prior Background

• Data Engineering or Data Science experience is non-negotiable , including:

  • Data pipelines / ETL / ELT / orchestration

  • ML or NLP model lifecycle

  • Analytics platforms or data product engineering

Good-to-Have / Preferred

• Fine-tuning and adaptation strategies (LoRA / PEFT / prompt tuning).

• Experience with MLOps / LLMOps platforms and observability stacks.

• Experience delivering GenAI solutions in regulated industries (Insurance, Healthcare, BFS).

• Exposure to enterprise AI governance frameworks .

Education

Bachelor’s or Master’s degree in Computer Science, Data Engineering, Data Science, AI/ML, or related fields (or equivalent practical experience).