Gen AI & Agentic AI Engineer

PwC Acceleration Center India; Bengaluru; On-site; INR 1,800,000 - 3,000,000; Full time

Job summary

PwC Acceleration Center India is seeking a hands-on GenAI engineer who builds AI solutions inside real client environments. Embedded within an enterprise client's team, you will develop LLM-powered apps, agents, and RAG pipelines from prototype to production-ready deployments.

You will write Python (and some TypeScript), design RAG pipelines with embeddings and vector stores, and collaborate with client teams via Teams, Slack, and email to deliver measurable business outcomes.

Qualifications

• Hands-on GenAI/LLM work in production or near-production settings.
• Experience building with LLMs on AWS Bedrock and provider APIs.
• Hands-on experience with at least one agent framework.
• Experience building RAG pipelines with embeddings and vector databases.
• Comfort with prompt engineering and model evaluation.
• Strong Python and some TypeScript, with API development skills.
• Data fundamentals across structured and unstructured data.
• Ability to work with ambiguity and a bias to build.

Responsibilities

• Build LLM-powered applications, agents, and automations for client problems.
• Design and implement RAG pipelines using embeddings and vector stores.
• Use agent frameworks to orchestrate multi-step workflows and tool use.
• Write clean Python (and some TypeScript) across backend and light frontend.
• Perform prompt engineering and evaluation to ensure quality.
• Integrate enterprise data into solutions and iterate with users.
• Collaborate with client teams via Teams, Slack, and email, documenting work.

Skills

GenAI/LLM work

Production-grade ML

Python

TypeScript

Agent frameworks

RAG pipelines

Embeddings/vector search

Prompt engineering

Client-facing collaboration

Tools

AWS Bedrock

OpenAI

Anthropic Claude

Pinecone

pgvector

Weaviate

Qdrant

OpenSearch

React

TypeScript

SQL/NoSQL

Docker

Kubernetes

CI/CD

Terraform

Databricks

Snowflake

Job description

A hands-on GenAI engineer who builds AI solutions inside real client environments. Embedded within an enterprise client's team, you will build LLM-powered apps, agents, and RAG pipelines that solve real problems, moving quickly from prototype to something the client can rely on, with AI tooling as your default way of working.

Key Responsibilities

• Build LLM-powered applications, agents, and automations that solve real client problems.
• Design and implement RAG pipelines using embeddings, vector search, and retrieval.
• Use agent frameworks to orchestrate multi-step workflows and tool use.
• Write clean Python (and some TypeScript) across backend and light frontend to ship end to end.
• Perform prompt engineering and set up evaluation to check the AI is doing its job.
• Integrate structured and unstructured enterprise data into solutions.
• Iterate quickly on feedback from real users.
• Use AI coding assistants and agent tooling as your default way of building.
• Collaborate with client teams over Teams, Slack, and email, and document your work.

Required Qualifications

• Software engineering experience with hands-on GenAI/LLM work in a production or near-production setting.
• Experience building with LLMs on AWS Bedrock and provider APIs (e.g., Anthropic Claude, OpenAI).
• Hands-on experience with at least one agent framework.
• Experience building RAG pipelines with embeddings, vector databases, and semantic search.
• Comfort with prompt engineering and basic model evaluation.
• Strong Python (and ideally some TypeScript), comfortable across backend and light frontend, building APIs.
• Good data fundamentals across structured and unstructured data.
• Comfort with ambiguity and a bias to build and learn.

Preferred Qualifications

• AWS AI services (Bedrock, SageMaker).
• Experience in regulated industries.
• Prior consulting, customer success, or forward-deployed experience.
• AWS Certified Solutions Architect Associate; an AI/ML certification such as AWS Certified Machine Learning Specialty.

Technical Skills & Tools

• Retrieval & RAG: embeddings, vector databases (Pinecone, pgvector, Weaviate, Qdrant, OpenSearch), semantic search
• Frontend: React, TypeScript
• Data: SQL / NoSQL, data pipelines, structured and unstructured data
• Good to have: AWS SageMaker, Docker, Kubernetes, CI/CD, Terraform, Databricks, Snowflake
• Comfort with ambiguity and a bias to build and learn.
• Clear communication with technical and non-technical audiences.
• Takes ownership of outcomes.
• Collaborative and fast-moving.

Experience Required

2-5 years.

Reporting & Team

Embedded within an enterprise client's team as part of our Forward Deployed Engineering practice, working alongside the client's engineers and our senior engineers.

Location & Work Model

• Work model: Forward-deployed and customer-facing; embedded within an enterprise client's team.
• Working hours: Overlap with client business hours (including US / EST); works at the client's delivery pace.