• Support development of LLM-powered agents (chatbot/voicebot) for task execution, retrieval, and simple memory.
• Implement and maintain RAG pipelines: chunking, embeddings, retrieval, reranking (with guidance).
• Build tool integrations: connect agents to internal APIs, databases, and third-party services.
• Assist with prompting / structured outputs and basic tool-calling workflows.
• Help productionize features: logging, evaluation datasets, basic telemetry/monitoring, bug fixes.
• Write clean, testable code and contribute to code reviews and documentation.
• Participate in experiments to improve response quality, latency, and reliability.

Must-Have Skills

• 0–2 years of experience (or strong projects) building AI/LLM applications.
• Solid Python fundamentals (APIs, packaging, basic testing).
• Familiar with using LLM APIs (OpenAI/Anthropic/HuggingFace or similar) for chat and structured outputs.
• Basic understanding of RAG and embeddings; exposure to a vector DB is a plus (FAISS/Qdrant/Pinecone/Weaviate/etc.).
• Comfortable working with REST APIs, JSON, Git, and backend basics.
• Good debugging mindset and willingness to learn production practices.

Nice to Have

• Exposure to agent frameworks (LangChain, LangGraph, CrewAI, AutoGen) or small agent projects.
• Experience with speech pipeline basics (STT/ASR, TTS) for voicebots.
• Familiarity with Docker and simple cloud deployment workflows.
• Built a demo UI with Streamlit/Gradio or basic web integration.

What You'll Learn / Growth Opportunities

• Agent patterns: tool contracts, memory/context handling, routing, multi-step workflows.
• Building reliable RAG systems and evaluation loops.