Agentic AI Developer / Engineer

We are seeking a skilled **Agentic AI Developer** to design, architect, and deploy intelligent, autonomous AI agent systems. In this role, you will go beyond traditional static LLM applications to build dynamic agents capable of multi\-step task decomposition, reasoning, tool execution, state management, and real\-time environment adaptation.

• *Key Responsibilities**

1\. Agent Architecture \& Orchestration

• Design and implement autonomous single\-agent and multi\-agent systems capable of multi\-step task planning, reasoning, and execution.
• Implement structured reasoning frameworks, such as Chain\-of\-Thought (CoT), ReAct, and Tree\-of\-Thoughts.
• Build short\-term and long\-term memory architectures using vector databases, key\-value stores, and graph\-based context engines.

2\. Integration \& Function Calling

• Connect AI models to external enterprise applications, APIs, databases, and third\-party tools via function calling, MCP (Model Context Protocol), and REST interfaces.
• Build dynamic retrieval systems (RAG, GraphRAG) to ground agent reasoning in private context and real\-world data.

3\. Safety, Guardrails \& Human\-in\-the\-Loop (HITL)

• Implement safety layers, output parsers, and validation frameworks to prevent agent hallucination, prompt injection attacks, and unintended actions.
• Architect Human\-in\-the\-Loop (HITL) approval gates for sensitive operational actions (e.g., financial execution, database modifications).

4\. Evaluation, Observability \& Performance

• Establish evaluation metrics for agent behavior, reasoning accuracy, tool use efficacy, and response latency.
• Utilize tracing tools (e.g., LangSmith, Phoenix, Datadog) to debug agent loops, optimize token/API usage costs, and identify execution bottlenecks.

• *Required Qualifications \& Skills**

Technical Core

• Programming: Advanced proficiency in Python (async/concurrency, type hinting, Pydantic) and modern API backend engines (FastAPI, Flask).
• Agent Frameworks: Experience with agent framework ecosystems like LangGraph, AutoGen, CrewAI, LlamaIndex, or Semantic Kernel.
• LLMs \& APIs: Deep understanding of foundational LLMs (OpenAI GPT, Anthropic Claude, Google Gemini, open\-source models like Llama) and fine\-tuning/prompt engineering tactics.
• Vector Databases \& RAG: Hands\-on experience with vector stores (Pinecone, Qdrant, Chroma, Weaviate, Pgvector) and semantic search architectures.
• Databases \& Tooling: Strong knowledge of SQL/NoSQL databases, Git, Docker, and REST/GraphQL APIs.

• *Soft Skills \& Mindset**

• System Design Thinking: Ability to decompose ambiguous human tasks into deterministic workflow loops.
• Product\-Focused: Focus on building safe, usable, and high\-utility features rather than speculative experimental code.

• *Preferred / Nice\-to\-Have Qualifications**

• Experience with MCP (Model Context Protocol) or building custom API tools for LLM agent integration.
• Familiarity with TypeScript / Node.js for client\-side or microservice integrations.
• Experience deploying containerized workflows on cloud platforms (AWS, GCP, Azure) with Kubernetes.
• Open\-source contributions to agentic AI libraries or AI evaluation frameworks.

• *What We Offer**

• Competitive salary and performance\-based equity/stock options.
• Hardware budget for local LLM prototyping (Apple Silicon, dedicated GPU access).
• Learning \& Development stipend for AI conferences, certifications, and research papers.

Pay: From ₹20,000\.00 per month

Benefits

• Flexible schedule

Experience

• Agentic AI: 1 year (Preferred)

Work Location: In person