Full-Stack AI Engineer – LLM Agents, MCP Servers & Applications

About the Role:
We are hiring a Full-Stack AI Engineer to design and build intelligent software agents using Large Language Models (LLMs). This is a high-impact, full-stack engineering role where you’ll build production-grade agentic applications across our core stack — FastAPI (Python), Next.js (TypeScript), and GCP (Google Cloud Platform) — with Cursor as your primary development environment.

You’ll be a key contributor in implementing MCP (Model Context Protocol) servers, which play a critical role in enabling rich context management and structured interaction between agents and models. Your work will span both backend infrastructure and frontend interfaces, powering scalable, context-aware AI systems.

Key Responsibilities:

• Design and build LLM-powered agent systems, including planning, tool use, memory, and user interaction flows.
• Implement and maintain MCP-compliant servers to facilitate structured, contextual communication between agents and models.
• Develop FastAPI-based APIs to orchestrate and serve intelligent agent workflows.
• Build dynamic, performant UIs in Next.js for interacting with agents and MCP-based systems.
• Use Cursor as your daily development environment for coding, collaboration, and debugging.
• Deploy and scale services on GCP, including Cloud Run, GKE, and Vertex AI.
• Implement robust multi-agent orchestration patterns using frameworks such as AutoGen, CrewAI, or LangGraph — enabling agents to collaborate, delegate, and reason over tasks in a coordinated manner.
• Integrate and manage vector databases, especially Qdrant, to provide long-term memory, context retrieval, and semantic search for LLM workflows.
• Collaborate closely with product and infrastructure teams to launch intelligent, user-facing tools.

Requirements:

• 3–7 years of full-stack software development experience.
• Strong experience with Python + FastAPI and TypeScript + Next.js.
• Hands-on experience building and deploying LLM applications using APIs or open-source models (OpenAI, Hugging Face, etc.).
• Familiarity with MCP concepts, such as structured prompts, shared model state, and protocol-driven interaction patterns.
• Experience designing or integrating MCP servers to support modular, context-rich LLM workflows.
• Practical knowledge of vector databases like Qdrant, including schema design and retrieval pipelines.
• Familiarity with orchestration frameworks (AutoGen, CrewAI, LangGraph) and multi-agent patterns (e.g., agent hierarchies, shared context).
• Experience deploying and managing services on GCP infrastructure, including CI/CD.
• Proficiency with Cursor as a development tool or a strong willingness to adopt it      fully.

Nice to Have:

• Experience with agent tool integration (e.g., search APIs, code execution environments).
• Understanding of prompt engineering, chaining, and dynamic task decomposition.
• Knowledge of LLM evaluation techniques, including test harnesses and human-in-the-loop review.
• Exposure to observability and debugging in distributed agent environments.
• Contributions to open-source agent or protocol-based systems, including MCP-compatible tools.

What You’ll Gain:

• A chance to build some of the most advanced, context-aware AI-powered agents in production.
• Full-stack ownership of intelligent systems that leverage the Model Context Protocol to improve reasoning, traceability, and modularity.
• A modern, fast-paced development experience using Cursor, backed by a collaborative and technically curious team.
• A front-row seat in shaping the next generation of LLM infrastructure and applications.