Senior AI Engineer

LearnWise AI is an AI-first scale-up building virtual assistants, instructor efficiency tools, and student engagement solutions for universities and colleges. We’re a small team moving fast: the best idea wins no matter who says it, and every hire owns real scope from day one.

About LearnWise

LearnWise AI is an AI-first scale-up building virtual assistants, instructor efficiency tools, and student engagement solutions for universities and colleges. We’re a small team moving fast: the best idea wins no matter who says it, and every hire owns real scope from day one.

About the role

LearnWise builds AI products for higher education that run inside Canvas, Brightspace, Moodle, and Blackboard. We are hiring a senior AI engineer who owns the systems underneath these products: agents that act safely on real systems, how we call and route models, where inference runs, and what it costs. You design it, write it down, ship it behind safe rollouts, and keep it reliable in production. This is production engineering around LLMs and agents, not model training.

What you’ll do

Build agents that act on real systems

• Design how agents call LMS and institutional APIs, with permissions, human approvals, sandboxed execution, and audit.

Build the AI platform

• Model routing, provider fallbacks, in-region inference, usage limits, and cost control.

Ship AI product features

• Take new learning, admin, and multichannel experiences from design to production.

Make AI observable and self-improving

• Detect silent failures, measure quality, and feed fixes back into prompts, workflows, and agent logic.

Extend the integration platform

• Grow the framework that connects SIS, CRM, ticketing, and other institutional systems.

Write designs before code

• Propose, align, then build; review others’ designs and code.

What you bring

• 6+ years as a software engineer building production backend services.
• Production experience in more than one backend language (e.g., Python, Go, TypeScript/Node, Java, Rust, PHP).
• Working Python knowledge and interest in learning async Python.
• LLM features shipped to production: tool-calling, structured outputs, streaming, RAG, embeddings, and their failure modes.
• Reliability under third-party dependency: retries, timeouts, fallbacks, rate limits, provider outages.
• You deploy what you build: containers, CI/CD, and a cloud platform (AWS preferred).
• Clear, direct written and spoken English; you raise blockers early.
• Daily use of AI coding tools, e.g., Codex or Claude, and you verify what they produce.
• Technical leadership: led development of a larger project or system across several engineers.
• Architecture design of services or AI platform components that others built on.

Nice to have

• AI Ops or agentic workflows: agents that call external APIs and take actions, with guardrails.
• LLM gateways or routers (LiteLLM or similar) and multi-provider setups (OpenAI, Anthropic, Bedrock, Azure OpenAI, Gemini).
• Retrieval at scale, including self-hosted embedding models.
• Observability and evals for AI systems (OpenTelemetry, Logfire or similar).
• Data residency and multi-region deployments (GDPR, FERPA).

• 6+ years as a software engineer building production backend services.
• Production experience in more than one backend language (e.g., Python, Go, TypeScript/Node, Java, Rust, PHP).
• Working Python knowledge and interest in learning async Python.
• LLM features shipped to production: tool-calling, structured outputs, streaming, RAG, embeddings, and their failure modes.
• Reliability under third-party dependency: retries, timeouts, fallbacks, rate limits, provider outages.
• You deploy what you build: containers, CI/CD, and a cloud platform (AWS preferred).
• Clear, direct written and spoken English; you raise blockers early.
• Daily use of AI coding tools, e.g., Codex or Claude, and you verify what they produce.
• Technical leadership: led development of a larger project or system across several engineers.
• Architecture design of services or AI platform components that others built on.

• AI Ops or agentic workflows: agents that call external APIs and take actions, with guardrails.
• LLM gateways or routers (LiteLLM or similar) and multi-provider setups (OpenAI, Anthropic, Bedrock, Azure OpenAI, Gemini).
• Retrieval at scale, including self-hosted embedding models.
• Observability and evals for AI systems (OpenTelemetry, Logfire or similar).
• Data residency and multi-region deployments (GDPR, FERPA).