DataArt

ml engineer for agent platforms

In short

Senior ML/AI platform engineer role focused on building an Agent Platform that lets AI agents discover, access, and use tools securely at scale, including MCP-based tool integration, semantic search, and policy-driven authorization. Requires 5+ years building AI/ML platforms or agentic systems, hands-on experience with Model Context Protocol, AWS Bedrock AgentCore Gateway, JWT-based cross-agent auth, and vector databases like FAISS or Pinecone. Fully remote, but limited to candidates based in Armenia, Bulgaria, Cyprus, Georgia, Kazakhstan, Latvia, Poland, Romania, Serbia, or Ukraine.

Описание: The Agent Platform enables AI agents to discover, access, and use tools securely and at scale. It supports protocol-based tool integration, agent-to-agent communication, semantic search, and policy-driven authorization for enterprise AI environments.
Задачи:
•Design tool integration layers for AI agents
•Implement Model Context Protocol or similar agent tool protocols
•Develop semantic tool discovery using embedding-based search
•Build authorization flows for tool calls and agent interactions using policy enforcement
•Implement JWT-based cross-agent authentication and identity federation
•Configure AWS Bedrock AgentCore Gateway for MCP target registration, HTTP target routing, and inference target configuration
•Collaborate on multi-agent orchestration and agent-to-agent communication patterns
•Support enterprise security reviews for AI systems, including authorization, data access, and audit
Требования:
•5+ Years of experience building AI or ML platforms or agentic systems
•Experience designing tool integration layers for AI agents, including function calling and tool use
•Experience implementing Model Context Protocol or similar agent tool protocols
•Experience with enterprise security reviews for AI systems, including authorization, data access, and audit
•Knowledge of agentic workflow patterns, including agent-to-tool communication, agent-to-agent communication, and multi-agent orchestration
•Experience with semantic tool discovery and embedding-based search
•Experience with agentic authorization and policy enforcement
•Experience with JWT-based cross-agent authentication and identity federation
•Experience with AWS Bedrock AgentCore Gateway
•Experience building multi-agent systems with agent-to-agent communication
•Experience with semantic search using embeddings, FAISS, Pinecone, or vector databases
Nice to have:
•Hands-on experience with AWS Bedrock Agents or AgentCore
•Experience with agentic gateway topology design including central and federated models and tool visibility scoping
•Experience with credential vending for agents including 3LO OAuth and per-tool authentication
Условия: Remote work is available in Armenia, Bulgaria, Cyprus, Georgia, Kazakhstan, Latvia, Poland, Romania, Serbia, and Ukraine.

Stop applying.
Get referred.

$10 in free credits

In just 60 seconds, fully on autopilot — Eraldt's AI agents read your CV, map the roles you fit, find live vacancies and the people who can refer you, write personal emails, and get you referred.

This isn't an application: your CV doesn't go to the company, and we don't promise a referral for this exact role.

Discover who else lands interviews on autopilot