Contact Us
Case Study  |  Digital Capability Building  |  Roche

Digital capability building, governed from day one.

A digital capability building case study: how we helped 80 Roche colleagues in Hong Kong learn to analyse, automate and build with Tableau, UiPath and JavaScript, and put AI governance in place before scale, not after it.

KY & Company Success Stories  |  Updated October 2026  |  7 minute read

Key takeaways

  • Digital capability building at Roche reached 80 staff in Hong Kong through a training academy, coaching and hackathons.
  • Three tool tracks covered analytics (Tableau), robotic process automation (UiPath) and programming (JavaScript).
  • An AI governance framework ran alongside: an AI policy, use-case risk assessment, approval workflow, AI register and AI training.
  • Running skills and governance together lets teams build quickly while every AI use stays accountable.

Digital capability building is now a priority for every life sciences organisation, because most have more data and more AI tools than people able to use them well. Business teams wait in an IT queue for a dashboard or an automation, while new AI tools arrive faster than anyone can check whether they are safe to use.

Our view is simple: digital capability is built by the people who do the work, and it lasts only when governance grows with it. So we ran both together: an academy and hackathons to build the skills, and an AI governance framework to keep what people build safe, compliant and accountable.

What the research says

Why digital capability building can’t wait.

The skills gap and the governance gap are growing together.

59 in 100

workers worldwide will need training by 2030, as 39% of core job skills are expected to change. Analytical thinking remains the top core skill, and AI and big data skills are growing fastest.

World Economic Forum, Future of Jobs Report 20251
88%

of organisations now use AI in at least one business function, yet only 7% say it has been fully scaled across the organisation.

McKinsey, The State of AI 20252
80%

of low-code tool users were forecast to be developers outside formal IT departments by 2026, up from 60% in 2021.

Gartner3
Employers name skill gaps as the single biggest barrier to business transformation, with 63% citing them as a major barrier to 2030. World Economic Forum, Future of Jobs Report 20251
Why it matters

Capability without governance is just more risk.

Every enterprise investing in digital capability building meets the same three pressures.

Business teams

The people closest to a process know best what to automate or measure. When only IT can build, good ideas wait in a backlog. Citizen developers are becoming the majority of low-code users,3 and they need real skills, not just a licence.

Leadership

AI use is almost universal, but scale is rare.2 Only 28% of organisations using AI say their CEO oversees AI governance,4 leaving accountability unclear just as adoption accelerates.

Compliance & data protection

Hong Kong’s Privacy Commissioner recommends an AI governance committee, risk assessment and staff training for any organisation using AI with personal data.5 In the EU, AI literacy for staff has been a legal duty since February 2025.6

The engagement at a glance

The digital capability building programme at a glance.

One programme, two outcomes: people who can build, and rules that keep it safe.

80

Roche colleagues in Hong Kong took part across business and operational teams.

Hong Kong
3

tool tracks: Tableau for analytics, UiPath for automation and JavaScript for programming.

Training academy, coaching & hackathons
5

AI governance components: policy, use-case risk assessment, approval workflow, AI register and training.

AI governance framework
Our approach

3 steps of digital capability building: learn, build, govern.

Three strands, run as one programme, so new skills turned into working solutions inside clear guardrails.

Step 1 · Learn

We teach the tools.

A structured academy taking colleagues from first login to confident use of analytics, automation and code.

Step 2 · Build

We coach real work.

Coaching and hackathons that turn learning into solutions for real business problems.

Step 3 · Govern

We make it safe to scale.

An AI governance framework so every new idea has a clear, fast path from proposal to approval.

Step 1 · Learn

We teach the tools.

Three tracks, each matched to a kind of problem.

Digital capability building training academy for Roche staff in Hong Kong
Digital capability building starts with hands-on learning on real business data.

The first step of digital capability building is teaching tools in the context of real problems. Training fails when it teaches software features instead of problems people recognise, so our academy was organised around three questions every team asks: what is happening in our data, what can we stop doing by hand, and what can we build ourselves?

Analytics with Tableau

  • Connecting to real data sources, cleaning data and designing dashboards people actually read.
  • Visual best practice: exceptions first, one message per chart, clear definitions for every metric.

Automation with UiPath

  • Spotting repetitive, rule-based tasks that are good robotic process automation candidates, and the ones that are not.
  • Building, testing and documenting robots so they can be handed over and maintained.

Programming with JavaScript

  • Core programming concepts for non-developers, applied to small tools and web-based workflows.
  • Writing readable, reviewable code that IT can support.

Training academy — Structured learning paths for 80 colleagues in Hong Kong, combining instructor-led sessions with hands-on exercises on realistic business scenarios. See our capability building services.

Step 2 · Build

We coach real work.

Skills stick when they solve a problem the same week.

Classroom skills fade quickly unless they are used. Coaching and hackathons gave colleagues a reason and a deadline to apply what they had learned to their own work.

  1. Pick a real problemTeams brought a process or a question from their own work, not a textbook exercise.
  2. Choose the right toolA dashboard, a robot or a small application, matched to the problem rather than the other way around.
  3. Build with a coachOur consultants coached teams from first idea to working prototype, unblocking them without taking over.
  4. Check it against the rulesAny idea using AI or sensitive data went through the governance path below before going further.
  5. Show it and hand it overHackathon teams presented their solutions, with documentation so good ideas could be supported and reused.

Coaching & hackathons — Time-boxed events and ongoing coaching that turned the academy’s skills into solutions, and surfaced the people ready to become champions for their teams.

Step 3 · Govern

We make it safe to scale.

An AI governance framework built for a regulated business.

Digital capability building creates more builders, so the organisation needs a simple answer to two questions: is this use of AI acceptable, and who decided? We designed an AI governance framework that answers both without slowing good ideas down. It reflects the recommendations of Hong Kong’s Privacy Commissioner5 and the structure of international frameworks such as the NIST AI Risk Management Framework7 and ISO/IEC 42001.8

Highlighted components are the five delivered in this engagement.

  • Proportionate, not bureaucraticLow-risk ideas move quickly; higher-risk uses get deeper review and clearer conditions.
  • One front doorEvery proposed AI use enters the same way, so nothing is adopted quietly.
  • Named ownersEvery approved use has an accountable owner, recorded in the register.
  • Humans stay accountableAI assists; people remain responsible for the decisions it informs.
  • Visible inventoryThe AI register shows what is in use, by whom, for what, and when it is next reviewed.
  • Literacy for everyoneTraining means users understand both what AI can do and where its limits are.

The same principles shape our own Responsible AI approach and the AI solutions we build for clients.

Before you scale

Six questions to ask before scaling digital capability building.

If you can’t answer these clearly, the skills you build may create risk faster than value.

  • Is training built around real business problems, or just software features?
  • Do people get coaching and a deadline to apply what they learned?
  • Is there one clear path for proposing a new AI or automation idea?
  • How is the risk of each AI use assessed, and who signs it off?
  • Do you know every AI tool in use today, who owns it and why?
  • Can IT support and maintain what citizen developers build?
FAQ

Digital capability building: frequently asked questions.

What is digital capability building?

Digital capability building is a structured programme that teaches employees to use data, automation and software tools to improve their own work. It usually combines training, coaching and practical projects, so new skills turn into working solutions rather than certificates.

What did the Roche digital capability building programme include?

The programme reached 80 Roche staff in Hong Kong through a training academy, coaching and hackathons, covering Tableau for analytics, UiPath for robotic process automation and JavaScript for programming. It ran alongside an AI governance framework.

What should an AI governance framework include?

A practical AI governance framework typically includes an AI policy, a risk assessment for each AI use case, an approval workflow with named decision-makers, a register of AI tools in use and training for staff. Hong Kong’s PCPD Model Framework and the NIST AI RMF describe similar elements.

Why combine AI governance with digital capability building?

Training more people to build with data and AI increases both value and risk. Putting governance in place from the start gives every new idea a clear path to approval, keeps AI use visible and accountable, and helps meet obligations such as AI literacy requirements.

Built for regulated environments

  • RBAC & Audit Trails
  • AES-256 Encryption
  • Selectable Data Residency — EU / US / SG
  • GDPR & CCPA Aligned
  • No customer data used for AI training
  • ISO 27001 Certification Targeted Feb 2027

References

  1. World Economic Forum (2025). The Future of Jobs Report 2025, published 7 January 2025.
  2. McKinsey & Company (2025). The state of AI in 2025: Agents, innovation, and transformation, 5 November 2025; see also AI at work but not at scale.
  3. Gartner (2022). Gartner Forecasts Worldwide Low-Code Development Technologies Market to Grow 20% in 2023, press release, 13 December 2022.
  4. McKinsey & Company (2025). The state of AI: How organizations are rewiring to capture value, March 2025.
  5. Office of the Privacy Commissioner for Personal Data, Hong Kong (2024). Artificial Intelligence: Model Personal Data Protection Framework, 11 June 2024.
  6. European Union (2024). Regulation (EU) 2024/1689 (Artificial Intelligence Act), Article 4 on AI literacy, applicable from 2 February 2025.
  7. National Institute of Standards and Technology (2023). Artificial Intelligence Risk Management Framework (AI RMF 1.0), NIST AI 100-1.
  8. International Organization for Standardization (2023). ISO/IEC 42001:2023 Artificial intelligence — Management system.

Digital capability building where it matters most.

Whether you want your teams analysing, automating or building with AI, the skills and the safeguards should grow together. Partner with KY & Company to design the academy, coach the work and put the governance in place.

Get In Touch →
Scroll to Top