Skip to content
Case Studies|4 min read

Building Technical AI Expertise

Claire Byrne · April 2026

Thanks to JSC for letting us share their story. JSC →

Back to Insights
Share

JSC is a New Zealand-owned manufacturer and supplier of premium architectural timber products, with sites across New Zealand and a strong focus on technical accuracy and compliance.

JSC brought Elements AI in to run a hands-on Claude training for the technical team, supporting technical use cases like documentation review and compliance preparation. The goal was to move the group from cautious, uneven AI usage to confident, practical workflows they could apply immediately in day-to-day work.

The starting point

Like many teams, JSC had a mix of AI experiences and tools across the group. Some people were already using AI for summarising and polishing writing, others were experimenting with ChatGPT or Copilot, and there was healthy skepticism based on real concerns:

  • Outputs that were long but not useful
  • Time lost to back-and-forth prompting to get to a usable result
  • Uncertainty about mistakes and the burden of review
  • Sensitivity around data security and what is safe to share with AI tools

What we did

Claire ran in-person, highly practical workshops focused on Claude, with demonstrations anchored in JSC's real work rather than generic examples. The session was structured to meet people where they were: we started by surfacing concerns, then worked through concrete techniques that reduce risk and improve output quality.

Key elements included:

  • Clarifying when Claude is the right tool and when it isn't, including a grounded comparison with ChatGPT and Copilot for common workplace tasks.
  • Teaching a repeatable prompting approach based on describing the full problem — defining audience, context, constraints, and the desired outcome.
  • Introducing workflow patterns that reduce errors and rework: planning before executing (so the team could validate an approach before generating long outputs); self-critique and confidence-checking to catch issues early; and using Projects and reusable Skills so good workflows don't need to be reinvented each time.

Company-specific setup

A key part of the workshop was showing the team how to make Claude reliably useful by giving it JSC-specific context and guardrails. Rather than relying on generic prompting, we covered how to set up Projects with the right reference documents and instructions — such as voice and tone, technical source material, and clear do/don't rules — so outputs stay consistent with JSC standards, reduce rework, and are safer to use in real client-facing and compliance-sensitive work.

Practical use cases from the session

Rather than focusing on AI in theory, the workshop surfaced and demoed immediate, role-relevant applications, including:

  • Drafting and improving responses to technical enquiries, including how to bring existing "canned answer" workflows into Claude.
  • Auditing technical documents against standards and revisions to reduce the risk of outdated references being circulated.
  • Turning dense information into clear summaries and decision-ready bullet points.
  • Producing leadership-ready outputs such as slide decks and structured documents without the usual overhead.

What the team said

"It's like working with an expert who has years of working with this market and knows what questions to ask."

Kate van Aarde, Marketing Manager — JSCKate van Aarde, Marketing Manager — JSC

"This is an enabler for us more than replacing us. It makes us do our jobs way better."

— JSC technical team

"The company is getting more value because we're able to use these skills to enhance what a typical person would be able to output."

Rubia Haiduk, Technical Manager — JSCRubia Haiduk, Technical Manager — JSC

What changed

By the end of the session, the team had:

  • A shared foundation for using AI safely, with risks named and addressed clearly.
  • Practical techniques to get higher-quality outputs faster, without requiring everyone to become a prompting specialist.
  • A clearer path from experimentation to adoption: start with a small set of high-value workflows, standardise them into reusable Skills, and build confidence through real usage.

What's next

This workshop was designed as the first step, not a one-off. Follow-up was planned to check adoption, capture what worked and what didn't, and shape the next phase around JSC's highest-value use cases and internal constraints.


Resources

Case StudyWorkshopTeam CapabilityHands-on AI
CB

Claire Byrne

Founder & Co-Chair of Elements AI. 25 years in product and technology. Now helping mid-market teams build real AI capability.

Related Insights

© 2026 Elements AI

Book Intro Call