← All case studies
Enterprise Software 400-person ML community built

Building AI Capabilities from Within

Client

A global mobile software group with over 1000 engineers across Asia and North America.

Challenge

The company faced rising costs and bottlenecks due to heavy reliance on a US-based ML team. Asia-based engineers, despite their scale, were sidelined from ML projects, leading to morale issues and underutilised potential.

Objective

Having recognised that the opportunity imbalance was stalling innovation and scalability, we suggested a plan to transform a disconnected engineering group into a self-sustaining ML capability in Asia — reducing external dependency, empowering local talent, and embedding AI deeply into the development cycle.

Approach

Partnering with the VP of Engineering, we helped reframe the situation from a talent gap to a strategic opportunity. We took a phased, high-leverage approach:

  • Articulated a bold, clear vision: to build a high-impact ML team of senior ICs (6–15 years of experience) focused on end-to-end product delivery, not just model prototyping.
  • Created a hiring plan to bring in senior ML engineers. Started weekly workshops to upskill local teams and build a shared ML library.
  • Directed the team toward emerging technologies — especially LLMs (back in 2022) — enabling early wins and future-proofing AI capabilities.
  • Built a culture where learning was the norm. Encouraged peer sharing, experimentation, and community-led development, culminating in a company-wide ML community with 400 participants.

Impact

  • Delivered several production ML projects quickly. This shifted perception: the Asia team was now seen as a core asset — not cheap labour.
  • Extended support to adjacent product groups, elevating the org’s overall AI fluency and reducing single-team bottlenecks.
  • Boosted morale, retention, and sense of ownership across the engineering group. Software engineers began proactively pitching ML ideas — a signal of deeper engagement.

Takeaways

  • AI transformation isn’t just about models — it’s about capability building.
  • The fastest route to ROI was not outsourcing ML, but growing internal leadership and ownership.
  • Betting early on key technologies helped the client leapfrog competitors while future-proofing their talent strategy.