Professional Certificate in AI Software Engineering
Graduate in AI Software Engineering, with the work to prove it.
Ship and operate real software through an AI agent, including platforms you have never used before.
Two twelve week semesters of real briefs, marked by someone who has done the job. You finish with a portfolio, a grade, and a certificate an employer can check for themselves.
$300
Semester 2
- 01Semester one, twelve weeksYou learn the workflows and use them on a brief every week. By the end you have a body of work rather than a folder of notes.
- 02Marked, not tickedCoursework and a final project come back with written feedback naming what to fix. This is the part that changes how you work.
- 03Semester two, twelve weeksYou pick a specialism and go deep enough to be hired for it, building the portfolio you will actually show people.
- 04The certificateIssued once both semesters are graded, carrying your grade, and verifiable by anyone you send it to. It can be failed, which is what makes it worth holding.
What is included
Everything the academy gives you is covered by what you pay: the videos, the guides, the downloads, the marking and your certificate. There is nothing further to buy from us.
You will use tools made by other companies, and some of them sell paid plans. You never need one to finish this programme. Every piece of work we ask for can be done on a free option, and where you need an account somewhere else we say so before that part starts.
If you do choose to pay for a tool, that is between you and them. The terms set this out in full.
Live support
Fewer learners means more attention, not a weaker cohort.
- Weekly build room or office hours for blockers and live walkthroughs
- Biweekly project critique focused on what is working, missing, and unsafe
- Capstone proposal review before the final build starts
- Final demo day where learners show the system, the tradeoffs, and the handover
- The live cadence stays in place even with two learners
- The instructor acts as the extra reviewer when peer volume is low
- Sample submissions are supplied when there are not enough peer projects to review
- Every learner keeps a visible progress thread in the community
Capstone proof
Revenue is not promised, but the final work must connect to a real person, workflow, or value case.
- A real user, real workflow, or real recurring problem behind the project
- A live demo, acceptance note, pilot feedback, or first outreach evidence
- A pricing, cost, or internal value model that shows why the work matters
- A handover record proving somebody else can understand and operate it
Who this is for
- You can prompt your way to a demo and cannot yet ship and run the result
- You want to operate platforms you have never learned, through an agent
- You want a credential an engineering manager can check
Before you start
What you should be able to do on day one. If any of these are shaky, say so when you enrol and we will point you at the remediation pack first.
- Clone a repository, run it locally, and push one branch before week one
- Set one environment variable and explain what would break if it was wrong
- Start one AI agent session and record the first useful correction you made
- Complete a short remediation pack before week one if setup or Git is weak
What you will be able to do
- Direct a coding agent through work you would have written yourself
- Connect the platforms you use so you can prompt them rather than click them
- Decide what an agent may reach, and bound what a mistake can cost
- Deploy, operate and recover a running system through an agent
- Ship work you can explain, revert and hand to somebody else
What is taught
10 modules across two semesters, in this order.
- 1Working inside an agent
- 2Asking for work
- 3Setting the agent up once
- 4Context engineering
- 5Connecting your tools: MCP, CLIs, GitHub, Vercel, Sheets, Linear, Slack
- 6Access and restrictions
- 7Version control with an agent
- 8Reviewing and testing what the agent wrote
- 9Deploying and operating with an agent
- 10Owning the result
What you will build
What you are holding at the end, not what you will have watched.
- A portfolio of marked engineering work
- A working system you built, connected and deployed through an agent
- A verifiable certificate
What a week looks like
What you hand in, every week, marked and returned with written feedback.
- Week 0: readiness check and setup proof
- Weeks 1-2: agent-directed bug fix and personal agent configuration
- Weeks 4-6: connected service, access model, and integration project
- Weeks 8-12: deployment recovery, peer review, and semester system capstone
- Semester 2: multi-agent work, custom MCP server, governed automation, measurement, and final handover
Where this applies
Work this skill is used for. Not a promise of employment.
- Engineering roles now asking for AI fluency
- Leading AI adoption inside an existing team
- Contract and product work building AI features
Who teaches it
Yerins Abraham
Co-Founder, Creovine Academy
Teaches this programme and marks the work himself. He builds the systems behind Creovine’s products and runs them in the company’s own businesses every day, including Lira, which handles real customer conversations in production, and Oystar, a clinical referral system used by hospitals in Rwanda. The briefs come from that work.
What you study
Every course in the catalogue is included, taught in this order.
What you walk away with
- A portfolio of finished work, made on real briefs rather than exercises
- A certificate an employer can verify, with your grade on it
- Every piece of work marked against a published standard, with your score
- A room of people learning the same thing at the same time, and the intake above you
- Every course in the catalogue, included for as long as the track runs