AI Intern at Brainwonders — Mumbai, India

About The Role

Type: Internship 6 months | Possibility of Full-time 
Focus Areas: 
-- Large Language Models (LLMs)
-- Agentic AI & Autonomous Systems
-- Multi-Agent Systems & Agent Orchestration
-- ML Fundamentals & Transformers
-- RAG, Embeddings & Vector Databases
-- Cloud Deployment & Scaling on AWS

Responsibilities

● Build and test LLM-powered applications, AI agents, and autonomous workflows. 
● Design and implement agentic AI systems capable of planning, reasoning, tool use, and task execution. 
● Develop multi-agent systems where multiple specialized agents collaborate to solve complex tasks.
● Experiment with agent frameworks such as LangGraph, CrewAI, LangChain, and other emerging agentic AI frameworks.
● Implement agent memory, state management, task decomposition, planning, routing, and orchestration. 
● Integrate agents with external tools, APIs, databases, search systems, and other services using function/tool calling. 
● Explore Model Context Protocol (MCP) and other approaches for connecting AI agents with external tools and data sources. 
● Use and adapt transformer-based models such as LLaMA, Mistral, and other open-source or proprietary LLMs for domain-specific applications. 
● Build RAG pipelines using embeddings and vector databases for knowledge retrieval and contextual reasoning. 
● Deploy AI models, agents, and supporting services on AWS (Lambda, EC2, S3, SageMaker, etc.). 
● Work on prompt engineering, structured outputs, fine-tuning, model selection, and pipeline optimization. 
● Develop evaluation strategies to measure agent accuracy, reliability, latency, cost, and task completion. 
● Implement monitoring, logging, and debugging for LLM and agentic workflows. 
● Collaborate with the team to prototype and productionize AI copilots, autonomous workflows, and intelligent automation systems.

Requirements

● Strong understanding of LLMs, tokenization, transformers, embeddings, and generative AI fundamentals. ● Good Python programming skills. 

● Familiarity with tools/libraries such as Pandas, NumPy, Hugging Face, LangChain, LangGraph, or CrewAI. ● Understanding of AI agent concepts, including tool calling, planning, reasoning, memory, state, and orchestration. 

● Basic understanding of multi-agent architectures and how multiple agents can collaborate or delegate tasks. 

● Familiarity with RAG, vector databases, semantic search, and embeddings. 

● Exposure to AWS services such as **EC2, Lambda, S3, ** 

● Basic knowledge of machine learning and hands-on project experience. 

● Ability to build and debug end-to-end AI applications. 

● Strong curiosity and willingness to experiment with rapidly evolving agentic AI technologies.

Good to Have 

● Hands-on experience building an AI agent or multi-agent system. 
● Experience with LangGraph, CrewAI, AutoGen, or similar agent frameworks. 
● Experience implementing tool/function calling and agent–API integrations. 
● Familiarity with MCP (Model Context Protocol). 
● Experience with agent memory, state management, planning, routing, and workflow orchestration. 
● Experience with Docker 
● Experience evaluating LLMs or agents using automated evaluation frameworks. 
● Contributions to open-source AI tools, models, agents, or frameworks. 
● Personal projects involving LLMs, RAG, AI agents, or multi-agent systems 
I nternship Details