I
Sr AI Engineer
Insight Global
2 hours ago
Full-time
On-site
Lincolnshire, Illinois, United States
Required Skills & Experience
7+ years professional software engineering, including significant production ownership (not just prototypes or coursework).\n3+ years building and shipping LLM, RAG, or agentic systems in production — or equivalent depth owning an internal AI platform.\nHands-on with LangChain and/or LangGraph (or comparable agent orchestration).\nExperience with vector databases / embeddings (pgvector, Pinecone, Weaviate, FAISS, OpenSearch, or similar).\nExperience building or integrating MCP servers / tool-calling interfaces for agents.\nStrong TypeScript and React / Next.js, plus backend in Python (FastAPI) and/or Node.js / Go.\nProduction cloud experience on AWS, GCP, and/or Azure; containers and CI/CD in real environments.\nPostgreSQL and at least one cloud data platform (Snowflake, BigQuery, or equivalent).\nOAuth 2.0 / SSO (Microsoft Entra ID preferred).\nProven ability to take AI-generated code and make it production-safe.\nComfortable in Git workflows, code review, and hybrid collaboration. 3 days in office required.
Nice to Have Skills & Experience
Daily use of Claude Code, Cursor, Lovable, or Copilot as a primary development workflow\nSupabase (Postgres + Auth + Edge Functions) and Vercel\nSnowflake MCP connectors or similar enterprise data-to-agent patterns\nKubernetes, Terraform, GitOps\nKong, Dynatrace, Datadog\nEvaluation harnesses, guardrails, prompt versioning, LLM observability (Langfuse, LangSmith, or similar)\nRetail, automotive, or dealership / field-ops technology
Job Description
We are seeking a Senior AI Engineer to own production AI systems end to end — not just wire prototypes. This role sits at the intersection of agentic AI, full-stack engineering, and AI platform / DevOps. You will design and ship LLM applications, RAG pipelines, MCP tooling, and the infrastructure that makes them reliable in an enterprise environment.\nYou will still work in our AI-assisted delivery flow (Lovable → GitHub → Claude Code → CI/CD → Vercel/GCP/Azure), but the bar is senior: you architect agent workflows, harden AI-generated code, own observability and deployment, and set patterns other engineers follow.\n\nKey Responsibilities\nAgentic AI, RAG & MCP\n\nDesign and ship production agentic systems using LangChain, LangGraph, and related orchestration patterns (supervisor/worker, tool-calling, human-in-the-loop, recoverable state).\nBuild and operate RAG pipelines: embeddings, hybrid retrieval, reranking, citation/grounding, and evaluation so answers stay accurate and auditable.\nStand up and maintain MCP servers and tool integrations so agents can safely call enterprise systems (Snowflake, GitHub, Slack, internal APIs, knowledge bases).\nWork with vector databases and embedding workloads (pgvector / Supabase, Pinecone, Weaviate, FAISS, or equivalent) for semantic search and agent memory.\nIntegrate multi-model LLM APIs (Claude, GPT, Gemini, Bedrock, etc.) with guardrails, cost/latency controls, and production observability.\n\nFull-Stack Production Engineering\n\nTake AI-generated React/Next.js (or equivalent TypeScript) front ends and turn them into secure, scalable full-stack applications.\nWire UIs to backends, APIs, PostgreSQL/Supabase, Snowflake, Salesforce, and internal microservices.\nImplement Entra ID (Azure AD) SSO, OAuth 2.0, and RBAC.\nDesign REST and GraphQL APIs with clear contracts, versioning, and enterprise auth.\n\nAI-Assisted Development (Senior Bar)\n\nUse Claude Code, Cursor, GitHub Copilot, and similar tools daily — and review, refactor, and harden AI-generated code before it ships (security, performance, maintainability).\nDefine team standards for vibe-coded prototypes moving through GitHub, Harness, and production.\nPartner with Agent Developers and AI Solution Architects to land agent capabilities in real applications, not demos.\n\nAI Platform, DevOps & Reliability\n\nOwn CI/CD (Harness, GitHub Actions, or equivalent) and deploy to Vercel, GCP (Cloud Run / GKE), Azure, and/or AWS.\nImplement observability with Dynatrace, Datadog, or equivalent (APM, logs, traces, alerting on both apps and LLM workflows).\nUse Docker/Kubernetes, IaC, and modern DevOps practices to keep AI services repeatable and recoverable.\nApply supply-chain and app security: secrets management, image scanning, input validation, prompt-injection prevention, and secure coding of AI-generated output.\nConfigure API gateways (Kong/Konnect or similar) for auth, rate limits, and traffic control.\n\npay rate: $60-90/hour depending on experience
7+ years professional software engineering, including significant production ownership (not just prototypes or coursework).\n3+ years building and shipping LLM, RAG, or agentic systems in production — or equivalent depth owning an internal AI platform.\nHands-on with LangChain and/or LangGraph (or comparable agent orchestration).\nExperience with vector databases / embeddings (pgvector, Pinecone, Weaviate, FAISS, OpenSearch, or similar).\nExperience building or integrating MCP servers / tool-calling interfaces for agents.\nStrong TypeScript and React / Next.js, plus backend in Python (FastAPI) and/or Node.js / Go.\nProduction cloud experience on AWS, GCP, and/or Azure; containers and CI/CD in real environments.\nPostgreSQL and at least one cloud data platform (Snowflake, BigQuery, or equivalent).\nOAuth 2.0 / SSO (Microsoft Entra ID preferred).\nProven ability to take AI-generated code and make it production-safe.\nComfortable in Git workflows, code review, and hybrid collaboration. 3 days in office required.
Nice to Have Skills & Experience
Daily use of Claude Code, Cursor, Lovable, or Copilot as a primary development workflow\nSupabase (Postgres + Auth + Edge Functions) and Vercel\nSnowflake MCP connectors or similar enterprise data-to-agent patterns\nKubernetes, Terraform, GitOps\nKong, Dynatrace, Datadog\nEvaluation harnesses, guardrails, prompt versioning, LLM observability (Langfuse, LangSmith, or similar)\nRetail, automotive, or dealership / field-ops technology
Job Description
We are seeking a Senior AI Engineer to own production AI systems end to end — not just wire prototypes. This role sits at the intersection of agentic AI, full-stack engineering, and AI platform / DevOps. You will design and ship LLM applications, RAG pipelines, MCP tooling, and the infrastructure that makes them reliable in an enterprise environment.\nYou will still work in our AI-assisted delivery flow (Lovable → GitHub → Claude Code → CI/CD → Vercel/GCP/Azure), but the bar is senior: you architect agent workflows, harden AI-generated code, own observability and deployment, and set patterns other engineers follow.\n\nKey Responsibilities\nAgentic AI, RAG & MCP\n\nDesign and ship production agentic systems using LangChain, LangGraph, and related orchestration patterns (supervisor/worker, tool-calling, human-in-the-loop, recoverable state).\nBuild and operate RAG pipelines: embeddings, hybrid retrieval, reranking, citation/grounding, and evaluation so answers stay accurate and auditable.\nStand up and maintain MCP servers and tool integrations so agents can safely call enterprise systems (Snowflake, GitHub, Slack, internal APIs, knowledge bases).\nWork with vector databases and embedding workloads (pgvector / Supabase, Pinecone, Weaviate, FAISS, or equivalent) for semantic search and agent memory.\nIntegrate multi-model LLM APIs (Claude, GPT, Gemini, Bedrock, etc.) with guardrails, cost/latency controls, and production observability.\n\nFull-Stack Production Engineering\n\nTake AI-generated React/Next.js (or equivalent TypeScript) front ends and turn them into secure, scalable full-stack applications.\nWire UIs to backends, APIs, PostgreSQL/Supabase, Snowflake, Salesforce, and internal microservices.\nImplement Entra ID (Azure AD) SSO, OAuth 2.0, and RBAC.\nDesign REST and GraphQL APIs with clear contracts, versioning, and enterprise auth.\n\nAI-Assisted Development (Senior Bar)\n\nUse Claude Code, Cursor, GitHub Copilot, and similar tools daily — and review, refactor, and harden AI-generated code before it ships (security, performance, maintainability).\nDefine team standards for vibe-coded prototypes moving through GitHub, Harness, and production.\nPartner with Agent Developers and AI Solution Architects to land agent capabilities in real applications, not demos.\n\nAI Platform, DevOps & Reliability\n\nOwn CI/CD (Harness, GitHub Actions, or equivalent) and deploy to Vercel, GCP (Cloud Run / GKE), Azure, and/or AWS.\nImplement observability with Dynatrace, Datadog, or equivalent (APM, logs, traces, alerting on both apps and LLM workflows).\nUse Docker/Kubernetes, IaC, and modern DevOps practices to keep AI services repeatable and recoverable.\nApply supply-chain and app security: secrets management, image scanning, input validation, prompt-injection prevention, and secure coding of AI-generated output.\nConfigure API gateways (Kong/Konnect or similar) for auth, rate limits, and traffic control.\n\npay rate: $60-90/hour depending on experience