Need AI Engineer - Agent Builder - Remote

Title: AI Engineer - Agent Builder

Location: Remote

Employment type Long Term Contract

Experience Minimum 6+ years in software engineering, including at least 1 or 2 year building LLM-based systems

About the role

We're hiring an AI Engineer who builds AI agents, not someone who mainly uses them.

Daily use of ChatGPT, Copilot, Claude or no-code AI tools is not enough for this role. You will design, code, deploy and maintain autonomous and semi-autonomous AI agents that do real work inside business workflows. Most of your time goes into agent logic, tool integrations, orchestration and evaluation pipelines, not into prompting a chat window.

If you've built an agent that calls tools, keeps memory, recovers from its own failures and runs in production, we want to talk to you.

What you'll build

AI agents built from the ground up. This covers planning and reasoning loops, tool and function calling, memory, and multi-step task execution.

Multi-agent systems. Specialized agents work together, hand off tasks and coordinate through an orchestrator.

Custom tools and integrations. These connect agents to APIs, databases, internal systems and third-party SaaS. The work includes building MCP (Model Context Protocol) servers.

RAG pipelines. This covers ingestion, chunking strategy, embeddings, vector search and retrieval tuning.

Evaluation frameworks. They measure accuracy, task completion, hallucination rate, latency and cost, and you use them to improve agents in a structured way.

Production deployments. Each one ships with logging, tracing, guardrails, human-in-the-loop checkpoints, error handling and cost controls.

Workflow automation with stakeholders. You turn manual business processes into automated agent solutions.

Must-have skills (builder level)

Strong Python at production quality, plus solid engineering fundamentals: APIs, async programming, testing and Git.

Hands-on agent building with LangGraph, LangChain, CrewAI, AutoGen, the OpenAI Agents SDK, the Claude Agent SDK, Semantic Kernel or LlamaIndex, or with custom agent loops written from scratch.

Direct use of LLM APIs (Anthropic, OpenAI, Gemini, Azure OpenAI, open-source models), including tool calling, structured outputs, streaming and token management.

Vector databases such as Pinecone, Weaviate, Qdrant, pgvector or Chroma.

Shipping AI to production with Docker, cloud deployment (AWS, Azure or Google Cloud Platform) and monitoring.

Understanding of agent failure modes, including looping, tool misuse, context overflow and prompt injection, and how to design around them.

Nice to have

Experience building MCP servers or custom tool ecosystems

Agent observability tools such as LangSmith, Langfuse, Arize or Weights & Biases

Fine-tuning or serving open-source models (Llama, Mistral, Qwen)

TypeScript/Node.js

Contributions to open-source AI or agent projects

What this role is NOT

This role is not a fit if your AI experience is mainly:

Using ChatGPT, Copilot or similar tools for your own productivity

Writing prompts without code around them

Configuring no-code or low-code automation tools (Zapier AI, Power Automate, etc.) without custom development

Completing AI certifications or courses without building and deploying anything

We value these skills, but this role requires engineering and shipping agent systems.

How to apply: show us what you've built

Along with your resume, include:

1. A link to at least one agent you built, such as a GitHub repo.

2. A 3 5 sentence write-up covering what the agent does, the tools or frameworks you used, the hardest problem you solved, and how you measured whether it worked.