Singapore’s Journey from Local Innovation to Global Impact with the AI Apprenticeship Programme (AIAP)

When AI Singapore started in 2017, we faced what seemed like an impossible challenge. We needed AI Engineers to work on AI projects, but we could not hire enough AI engineers. This article describes the AI Apprenticeship Programme (AIAP) and why it is putting a mark on the global stage.

Traditional academic routes were not producing industry-ready talents fast enough, and expensive vendor certifications were not creating the practical, real-world skills needed.

In my earlier article From Supercomputers to LLMs: How Open Source and Ingenuity Democratize Cutting-Edge Technology, I shared how constraints sometimes generate innovations. Similarly, constrained by the limited budget and “unglamourous” nature of a university programme, I found it hard to entice experienced Singaporean AI engineers to join AI Singapore, hence I need an alternative source of finding and hiring these engineers.

We needed to innovate. We needed a blue ocean strategy. We not only solved our AI talent problem for ourselves, but also created a model that is now being studied and copied worldwide.

The AIAP Way: Growing Our Own Timber

Instead of fighting in the “red ocean” of competing for limited AI talents, we created our own “blue ocean” of AI Talents.

Apprenticeship Model

I created the AI Apprenticeship Programme (AIAP) to identify passionate Singaporeans keen to deepen their current AI/ML/Data-Science skills and move from learner to a real-world practising AI engineer. Features of AIAP:

    • Is a 9-month, full-time, in-person intensive programme with programme fees fully subsidized by the Singapore government, including a monthly stipend of $3,500 – $5,500 per month.
    • 3-months of deep-skilling based on self-directed learning, group projects and discussions, code-walkthrough. No lectures!
    • 6-months to develop an MVP that typically comprises an end-to-end AI/ML processing pipeline, model building, training and deployment. These are real-world customer projects which the customer co-invest $150,000 to get the MVP done. They are not toy internship projects with inconsequential business outcome.
    • Expect to work with missing and messy data and changing customer requirements.
    • Mentorship from AI engineers whom have gone through the same journey as you!

    One of the biggest misconceptions that hiring managers have is that only Computer Science graduates can do AI and ML. Some of my best AI engineers and data scientists have degrees in mathematics, economics, psychology, business, biology or industrial engineering. What we seek is the ability to learn fast, passion for solving data problems, and a love of working with data. Having a STEM degree is good but is not a requirement to get into the AIAP programme. In fact, a degree isn’t even necessary! You just need to pass our technical assessment and interviews!

    Win-Win-Win Industry Integration

    Another unique aspect of the AIAP is that the apprentices deliver a real-world MVP for a customer project.

    • It is a win for the customer as they get their business use case solved.
    • It is a win for the apprentice as they to horn their skills on a real-world project.
    • It is a win for Singapore (AI Singapore), as we get to enable another organization on their AI journey, and another 3-6 Singaporeans trained to become an AI Engineer.

    We forged our AI Apprenticeship Programme (AIAP) in the real world of customers’ AI business use case, working with them to ideate, co-create, train and deploy over 200 AI and LLM projects to date.

    Why Traditional Training Methods Does Not Work!

    Let me be direct here – 6-weeks (or even 6-months) bootcamps claiming to train you to be a data scientist even if you do not know Python is just fake news! The reality is:

    Academic Programs: While universities do excellent research and teach the foundations, their graduates often lack practical experience with state-of-art open-source tooling for development work, real-world messy data and changing customer requirements.

    Vendor Certifications: Only good to display on your LinkedIn profile. We have encountered many hopeful candidates with impressive professional vendor certifications in AI, ML and Data Science, and yet they fail in our basic AIAP technical assessment and interviews.

    Short Bootcamps: 5-day coding bootcamps promising to land you a job is just wrong. You need months, if not years, of training and hands-on work to become proficient.

    Hackathons: Frequently organized by corporate innovation teams, their purpose is not always about tangible results but fulfilling internal KPIs, such as a yearly quota of hackathons, number of teams/people participation – all about outreach and branding. Hackathons impart little knowledge, build limited skills, and rarely result in viable outcomes.

    From Local to Global

    What started as a solution for Singapore’s AI talent gap has evolved into a model that countries worldwide are adopting. Here’s why:

    1. Speed to Market

    • Our AIAP+100E model creates deployable AI engineers in 9 months
    • Real projects mean real experience, not just theoretical knowledge

    2. Industry-Ready AI Talents

    • Countries are realizing what we discovered – you can build AI talent from diverse backgrounds.
    • An extra benefit is that these AI Engineers comes with domain expertise from the industry

    Global Adoption Stories

    Let me share some specific examples of how different countries are implementing our model:

    1. Egypt’s Experience

    • Adopted our AI For Everyone, 100E and AIAP model with local customization
    • Focusing on their strengths in mathematics and engineering

    2. Serbia’s Journey

    • Close collaboration with AISG to implement the AI Readiness Index for their SMEs
    • Creating their own version of 100E and AIAP suited to local industry needs

    3. Isle of Man’s Digital Goal

    • Train everyone to be AI-Aware
    • Identify and groom local technology partners to accelerate their company’s adoption of AI solutions.

    4. Rwanda’s Vision

    • Building Africa’s first AI-ready workforce
    • Focus on developing solutions for healthcare and agriculture
    • Creating a hub for AI talent in Africa

    5. European Union’s Interest

    • Studying our governance frameworks and talent development model
    • Particularly interested in our AI Readiness Index (AIRI) and AI Apprenticeship Programme (AIAP)
    • Looking at adapting our approach for the European context

    What Makes Our Model Attractive?

    1. Proven Track Record

    • Over 400 AI engineers trained and employed
    • Over 200,000 Singaporeans made AI-Aware
    • Over 200 real-world AI projects deployed

    2. Comprehensive Ecosystem Approach
    We do not just train engineers. Our model includes:

    The Future: Building an AI-First World

    As we look ahead, we are seeing an exciting transformation:

    • More countries adopting and adapting our model
    • Regional AI talent hubs emerging
    • Cross-border collaboration increasing
    • A growing global network of AI-ready nations

    Remember this: Everyone can learn to program, but not everyone has a passion for data and solving real-world data problems. Focus on finding and nurturing those who do. The global adoption of our model proves something I have always believed – successful AI transformation is not about having the biggest budgets or the most advanced technology. It is about having the right framework to develop talent and deploy solutions that create real impact.

    As we continue to share our experience with other nations, we are not just building an AI-First Nation – we are helping create an AI-First World. And we are just getting started.

    Leave a Reply

    Your email address will not be published. Required fields are marked *