Agentic Application Engineer

EY; Pune District; On-site; INR 1,200,000 - 1,800,000; Full time

Job summary

EY is seeking a Senior Agentic Application Engineer to design and scale agentic AI systems with emphasis on core algorithms, model behaviour, and architecture. You will own end-to-end development from reasoning to deployment within enterprise settings.

The role requires 5-9 years of experience, hands-on expertise in agentic frameworks, LLM internals, and modern AI tooling. Onsite Pune location with 5-day work weeks.

Qualifications

• 5-9 years of experience building AI-powered applications with ownership of algorithms and architecture.
• Strong understanding of agentic systems, autonomous workflows, and multi-agent architectures.
• Ability to design agent frameworks from scratch or significantly customise existing ones.
• Deep knowledge of LLM internals, transformer architectures, tokenisation, and inference optimisation.
• Experience with open-source LLMs such as Llama, Mistral, Mixtral, Falcon, Qwen.
• Proficient in Python and/or Java/Go for backend AI systems.

Responsibilities

• Design, architect, and develop enterprise-grade agentic AI systems with ownership of model behaviour and execution workflows.
• Build custom agent orchestration frameworks, memory management, and stateful workflows.
• Develop and optimise reasoning, planning, tool-use, and fallback strategies for autonomous agents.
• Implement and fine-tune open-source and self-hosted LLMs with cost and security considerations.
• Design advanced RAG architectures, embedding pipelines, and knowledge-grounding solutions.
• Integrate AI agents with enterprise apps, APIs, data platforms, and ML ecosystems.
• Establish reusable design patterns, libraries, and internal frameworks for agentic development.
• Optimise model inference performance, latency, and cost through programmatic reasoning.
• Implement Responsible AI controls including governance and guardrails.
• Collaborate with stakeholders to translate requirements into reliable AI solutions.
• Lead production support, troubleshooting, and performance optimisation across system and model layers.
• Mentor junior engineers on AI fundamentals and best practices.

Skills

Agentic systems

LLM internals

LangChain

Python

GCP deployment

MLOps

AI security

Java/Go

insurance domain knowledge

Education

BE/B.Tech or equivalent

Tools

LangChain

LangGraph

AutoGen

CrewAI

Llama

Mistral

Mixtral

Falcon

Qwen

Job description

Senior Agentic Application Engineer

Experience - 5 to 9 years

Notice Period - Immediate to 60 Days

Location: Pune (Onsite) 5 days

Role Overview:

We are seeking a highly skilled Agentic Application Engineer to design, develop, and scale intelligent agent-based systems with a strong focus on core algorithms, model behaviour, reasoning frameworks, and system architecture. The ideal candidate will have deep expertise in building agentic solutions from first principles, with minimal reliance on abstraction-heavy orchestration frameworks and managed AI services.

Key Responsibilities:

• Design, architect, and develop enterprise-grade agentic AI systems with ownership of model behaviour, decision-making logic, and execution workflows.
• Build custom agent orchestration frameworks, memory management systems, context-handling mechanisms, and stateful workflows.
• Develop and optimise reasoning, planning, tool-use, fallback, and error-handling strategies for autonomous agents.
• Implement and fine-tune open-source and self-hosted LLMs, selecting optimal deployment strategies based on cost, performance, and security requirements.
• Design advanced RAG architectures, embedding pipelines, retrieval mechanisms, reranking models, and knowledge-grounding solutions.
• Integrate AI agents with enterprise applications, APIs, data platforms, and internal machine learning ecosystems.
• Establish reusable design patterns, libraries, and internal frameworks for agentic application development.
• Optimise model inference performance, latency, cost efficiency, and response quality through code-driven and programmatic reasoning techniques.
• Implement Responsible AI controls including explainability, governance, auditability, and guardrail mechanisms.
• Collaborate with business and technical stakeholders to translate complex requirements into reliable and scalable AI solutions.
• Lead production support, troubleshooting, root cause analysis, and performance optimisation across both system and model layers.
• Mentor junior engineers and consultants on AI fundamentals, system design, algorithms, and best practices.

Education:

• Bachelor's degree in computer science, Engineering, or a related discipline (BE/B.Tech or equivalent).

Required Experience & Technical Skills:

• 5-9 years of experience building AI-powered applications with hands-on ownership of algorithms, model behaviour, and architecture.
• Strong understanding of agentic systems, autonomous workflows, multi-agent architectures, and custom orchestration techniques.
• Ability to design agent frameworks from scratch or significantly customise existing frameworks.
• Deep knowledge of LLM internals, transformer architectures, tokenisation, attention mechanisms, inference optimisation, and model trade-offs.
• Experience working with open-source LLMs such as Llama, Mistral, Mixtral, Falcon, Qwen, or similar models.
• Expertise in advanced reasoning approaches including planners, graph/tree-based reasoning, critic-agent loops, and tool-aware agents.
• Practical experience with frameworks such as LangChain, LangGraph, AutoGen, and CrewAI, along with the ability to customise or replace them when required.
• Strong preference for code-first, algorithm-driven development over black-box AI implementations.
• Deep expertise in RAG systems, vector search, embeddings, hybrid retrieval, reranking, indexing, and information retrieval techniques.
• Strong programming expertise in Python and/or Java/Go for backend AI systems.
• Experience deploying AI workloads on GCP, including GPU infrastructure, containers, scalable inference services, and self-managed deployments.
• Good understanding of MLOps practices including model versioning, experimentation, evaluation pipelines, and controlled rollouts.
• Hands-on experience with AI observability, reasoning traceability, latency analysis, quality evaluation, and cost monitoring.
• Strong knowledge of AI security, privacy, access controls, compliance, and enterprise governance frameworks.
• Prior experience within the General Insurance domain (Health, Motor, or Travel Insurance) is preferred.

Soft Skills:

• Strong ownership mindset with a passion for solving complex technical challenges.
• Ability to thrive in ambiguous environments and drive solutions independently.
• Excellent collaboration skills across engineering, data science, and business teams.
• Strong communication skills with the ability to articulate technical decisions and trade-offs to both technical and non-technical stakeholders.
• Demonstrates professionalism, accountability, and a commitment to delivering high-quality outcomes.