Ignite Builder

Typical client situations

Examples of when the Ignite Builder profile is the right one for the client.

• 01The client has an AI idea that works as a prototype but needs to be in actual production with users
• 02The client needs to build an agent that performs tasks (not just answers) against internal systems and data
• 03The client must replace a manual step with something that handles edge cases and failures gracefully
• 04The client wants to connect AI to their own systems via APIs, MCP or custom integrations

Core skills

• LLM app engineering: structured outputs, tool use, streaming, multi-turn state
• Agent orchestration (LangGraph, Pydantic AI, multi-agent patterns)
• Model Context Protocol (MCP) as integration layer to enterprise data and actions
• RAG engineering: chunking, hybrid search, re-ranking, retrieval evaluation
• Eval CI: regression testing of prompts, golden sets in git, LLM-as-judge
• Cost engineering: prompt cache, semantic cache, model router, batching
• Secure integration: authentication, secret management, audit logging

Disse kursene er unike for Ignite Builder. Felleskursene som alle Ignitere tar finner du på kursoversikten.

• Builder

  In Discover

  Provider-independent LLM architecture

  Why you should speak three dialects of LLM.

• Builder

  In Build

  Robust AI services in production

  Streaming, retries, circuit breakers — patterns that keep agents alive in production.

• Builder

  In Build

  Quality gates in the delivery pipeline

  Regression-test prompts and agents — break the build when quality drops.

• Builder

  In Operate

  Azure deployment for Norwegian clients

  Cloud architecture that satisfies Norwegian data requirements — and that you can defend in a security committee.

• Builder

  In Operate

  MCP integrations and custom server design

  Connect agents to enterprise data and actions through Model Context Protocol.

• Builder

  In Operate

  AI deployment and infrastructure as code

  From notebook to production: build the infrastructure your AI solution actually lives in.

• Builder

  In Operate

  Agent deployment and harness

  From a working agent to an agent that survives production.

• Builder

  In Adopt

  User feedback and instrumentation

  Build AI systems that learn how they're actually used.

Certifications

• Claude Certified Developer – Foundations (required): Anthropic's developer certification. Documents that you build correctly against the Claude platform — API, tool use, MCP and agent patterns. The curriculum is covered through the Build phase; the exam is 120 minutes of scenario-based questions at Pearson VUE.
• Microsoft AI-102 – Azure AI Engineer Associate (required): Documents that you can run AI solutions in production on Azure, the platform many of our clients are on. Prepared through the cloud and operations courses.
• Claude Certified Architect – Foundations (recommended next step): For those who want to design end-to-end agent architectures. A natural continuation after graduation.