2 days ago
Bengaluru, IndiaMid Level / Senior
Responsibilities
- Design and develop autonomous and semi-autonomous AI agents with multi-step reasoning, planning, decision-making, defined roles, inter-agent communication, and collaborative execution.
- Implement agent orchestration using LangGraph, AutoGen, CrewAI, or custom frameworks, including short-term, long-term, and episodic memory.
- Integrate LLM providers and APIs, design prompts and system instructions, implement RAG and function-calling pipelines, and validate and parse model outputs.
- Build FastAPI-based microservices, REST and streaming APIs, event-driven triggers, task queues, scheduling, asynchronous execution, state management, and session persistence.
- Create agent tools for file processing, web search, SQL querying, SFTP, third-party APIs, MCP, and custom tool registries.
- Connect agents to structured and unstructured data sources and vector databases for semantic retrieval.
- Implement guardrails, output validation, human-in-the-loop checkpoints, sandboxing, error handling, retries, and graceful degradation.
- Instrument, evaluate, and monitor agents for accuracy, latency, tool-use correctness, goal completion, hallucination, escalation, and auditability.
Requirements
- 4–7 years of professional experience.
- Bachelor of Technology or equivalent listed qualification such as BE/BTech, MCA, MTech, or MBA.
- Advanced Python programming, including asynchronous programming, object-oriented programming, and design patterns.
- Experience with LangChain, LangGraph, AutoGen, CrewAI, or equivalent agent frameworks and LLM APIs.
- Experience with OpenAI API, Anthropic Claude, HuggingFace Transformers API, prompt engineering, chain-of-thought, ReAct, structured outputs, and function calling.
- Experience developing FastAPI services and REST or streaming APIs using SSE or WebSockets.
- Experience with Pandas, structured and unstructured data parsing, Redis, Kafka, RabbitMQ, and PostgreSQL.
- Experience with logging, tracing, metrics, and observability tools such as LangSmith or OpenTelemetry.
- Preferred qualifications include fine-tuning or RLHF pipelines, MCP, Ollama, vLLM, LiteLLM, Kubernetes, data engineering or ingestion platforms, AI safety, responsible AI, model risk management, and browser-use or computer-use agents.
Benefits
- PwC offers inclusive benefits, flexibility programmes, wellbeing support, mentorship, and opportunities for learning and career growth.
- The role is within PwC Advisory in Bangalore; no specific remote, hybrid, travel, or contract details are stated.
Tech Stack
Categories
About PwC
At PwC, we help clients drive their companies to the leading edge. We’re a tech-forward, people-empowered network with more than 370,000 people in 149 countries. Across audit and assurance, tax and legal, deals and consulting we help build, accelerate and sustain momentum. Find out more at www.pwc.com. PwC: Audit and assurance, consulting and tax services PwC refers to the PwC network and/or one or more of its member firms, each of which is a separate legal entity. Content on this page has been prepared for general information only and is not intended to be relied upon as accounting, tax or professional advice. Please reach out to your advisors for specific advice.