Engineering: Agentic AI Engineer

Python, TypeScript, OOP, DSA, Git, Linux, REST APIs, and testing.

ML, deep learning, neural networks, NLP, transformers, embeddings, and fine-tuning.

LLM architecture, API integration, prompt and context engineering, structured outputs, tool calling, and cost/latency optimization.

Retrieval-augmented generation, vector databases, hybrid search, reranking, chunking, and knowledge graphs.

Agent architecture, planning, reasoning, memory, tool use, self-correction, multi-agent systems, and orchestration.

Deep work in LangGraph with hands-on exposure to LangChain, CrewAI, AutoGen, OpenAI Agents SDK, ADK, and LlamaIndex.

MCP servers, clients, tools, resources, prompts, auth, and connecting agents to external applications.

n8n, Make, Zapier, webhooks, event-driven and browser/database/email/CRM automation for business processes.

SQL (PostgreSQL, MySQL), NoSQL (MongoDB, Redis), and vector stores (Qdrant, Pinecone, Weaviate, FAISS).

FastAPI, Node.js, REST/GraphQL, WebSockets, auth (OAuth/JWT), microservices, jobs, and queues.

AWS, Azure, or GCP, plus Docker, Kubernetes, CI/CD, GitHub Actions, Terraform, and serverless.

Model and prompt versioning, experiment tracking, monitoring, tracing, logging, and CI/CD for AI.

LLM, agent and RAG evaluation, hallucination detection, groundedness, red teaming, and evaluation metrics.

Prompt injection, jailbreaks, agent hijacking, permissions, secrets management, guardrails, and AI safety.

Logging, tracing, metrics, agent trajectories, token/cost/latency monitoring, and distributed tracing.

Vision, speech-to-text, image/document understanding, OCR, and multimodal agents.

Multi-agent and hierarchical systems, supervisor and planner-executor architectures, ReAct, and stateful long-running agents.

Requirement gathering, automation analysis, AI ROI, system design, product thinking, and client communication.