Saffa SkillsSaffa Education & Training
Menu

Technology pathway

Software Developer to AI Engineer

Build on solid development experience to design, test and ship AI-powered features that are reliable, safe and worth using.

Pathway status: Coming later7 stages

Who it’s for: Working software developers who want to build reliable, safe applications on top of modern AI models.

You are hereSoftware Developer

7 stages

TargetAI Engineer

Starting point

What you bring, and what to build

Developers already write, test and maintain software, work with APIs and debug complex systems. AI engineering uses all of that, with new challenges: models that do not always give the same answer, retrieval of the right information, evaluation and responsible use. This pathway builds those skills and ends with a practical AI application you design, test and explain.

Transferable strengths you already have

  • Programming experience
  • Working with APIs
  • Testing and debugging
  • Version control
  • System design thinking
  • Reading technical documentation

Capability areas to build

Skills to build

  • How language models work
  • Working with AI APIs
  • Prompt and context design
  • Retrieval from documents and data
  • Tool use and agents
  • Evaluation and testing
  • Cost and performance trade-offs
  • Responsible AI
  • Protecting user data

The route

Stage by stage

Each stage shows its own status. None of these stages’ training is being built yet; the pathway as a whole is coming later.

You are here

Software Developer

  1. Stage 1

    Status: Coming later

    How modern AI models work

    Understand enough about how models work to predict their strengths, limits and failure modes.

    • How language models work
    • Limits and failure modes

    Builds: Knowledge & Theory

    Training: AI & Automation, Software Development

  2. Stage 2

    Status: Coming later

    Working with AI APIs

    Call models from your code, design prompts and context, and handle errors properly.

    • Working with AI APIs
    • Prompt and context design
    • Error handling

    Builds: Practical / Occupational Skills

    Training: Software Development

  5. Stage 5

    Status: Coming later

    Evaluation and testing

    Measure whether an AI feature actually works, and catch regressions before users do.

    • Evaluation and testing
    • Cost and performance trade-offs

    Builds: Practical / Occupational Skills

    Training: Software Development

  6. Stage 6

    Status: Coming later

    Safety, privacy and responsible AI

    Protect user data, defend against misuse and design for fairness and transparency.

    • Responsible AI
    • Protecting user data
    • Prompt injection awareness

    Builds: AI & Modern Technology

    Training: AI & Automation, Cybersecurity

  7. Stage 7

    Status: Coming later

    Practical project

    Design, build, evaluate and explain a small AI application that solves a real problem.

    • Applying AI engineering skills
    • Explaining your design decisions

    Builds: Practical / Occupational Skills

    Training: Software Development

Target

AI Engineer

Practical evidence

What you’ll be able to show

Progress is proven with real work, recorded in your private Skills Passport.

  • A working AI application in a code repository
  • An evaluation set with results and notes
  • A short design document explaining trade-offs
  • A written review of privacy and misuse risks

Linked training

Training areas used

Where next

Possible next pathways

  • Status: Coming later

    FromIT Support to ToCybersecurity Analyst

    Build on hands-on IT support experience with security fundamentals, networks, identity, monitoring and incident response to move into security work.

    7 stages

All pathways

Launch updates

Hear when this pathway opens

Pathways open in stages as their training is built. Join the launch list to hear first.

Get notified