AI Value Discovery Day: Broomfield Construction

Client: Broomfield Construction

Event: GEFIE 2 – AI Value Discovery Day

Industry: Construction

Date: 12th February 2026

Consultant: Rufus Curnow

Applying AI to Tendering, Invoicing and Operational Visibility in Construction

Broomfield Construction partnered with Dixon AI to deliver a GEFIE 2 AI Value Discovery Day in Ireland, focused on turning existing AI literacy into practical, workflow-level application. Building on previous learning, the session moved from prompting fundamentals to reusable tools, custom bots and structured innovation. The outcome was a defined portfolio of AI ideas progressing towards hackathon development, aligned to the AI Transformation Playbook’s Formulate stage, where experimentation becomes purposeful and value-led.

Broomfield Construction operates in a sector where margins, timelines and coordination are tightly managed. In construction, improvements in tendering accuracy, invoice handling or delay analysis have direct commercial impact. Following earlier literacy development, the organisation had reached the point where awareness needed to translate into applied experimentation.

Within the AI Transformation Playbook, this corresponds to Stage 4: AI Application and Experimentation. The objective is not abstract exploration, but structured use of AI within real roles, processes and commercial constraints.

Objectives of the Event

The AI Value Discovery Day was designed to:

• Consolidate prompting techniques introduced in earlier sessions

• Transition from one-off prompts to reusable, role-specific assets

• Introduce and build custom GPTs and clarify the distinction between custom GPTs and Projects

• Identify practical AI applications across core construction workflows

• Generate a defined set of innovation ideas to progress into the upcoming hackathon

What Happened During the Event

The day began by revisiting prompting techniques previously introduced, reinforcing structure and clarity. Rather than introducing entirely new concepts, the focus was on consolidation and refinement. Participants strengthened their ability to frame requests clearly and maintain purpose within an interaction.

From there, the session moved into reusable prompts. Each participant built at least one prompt tailored to their role. The emphasis was on creating assets that could be saved, reused and refined rather than recreated from scratch. This marked a shift from experimentation to operational efficiency.

The group then moved into bot creation. Participants built custom GPTs and explored how these differ from Projects, gaining practical understanding of when each structure is appropriate. This hands-on build phase anchored AI capability within real responsibilities rather than abstract demonstrations.

In the afternoon, the session transitioned into a structured innovation workshop. Participants identified operational friction points and explored how AI could address them. Ideas were expanded collaboratively, with a view to progressing the strongest concepts into the next-stage hackathon.

The level of engagement was high throughout. Questions were practical, grounded in live construction challenges, and focused on how tools could support existing workflows.

Insights Emerging from the Session

Several themes became clear during the day:

First, reusable assets matter. The shift from single prompts to structured, saved prompts and custom GPTs changes how AI integrates into daily work. When individuals stop starting from scratch, AI moves from novelty to infrastructure.

Second, construction presents significant opportunities in information flow. Ideas generated around invoicing, tendering and delay analysis reflected a shared recognition that large volumes of documentation and data create both complexity and opportunity.

Third, visibility of assets and people remains a live operational challenge. Proposed solutions included logging systems to understand the location of plant and equipment, as well as tracking where engineers are deployed. These concepts demonstrate how AI can support coordination and oversight in project-driven environments.

Finally, collaborative idea generation accelerates confidence. When participants see peers building and applying tools within similar roles, experimentation becomes normalised.

Impact

By the end of the session, Broomfield Construction had:

• Strengthened prompting discipline and clarity

• Created reusable prompts tailored to individual roles

• Built custom GPTs aligned to specific workflow needs

• Generated a portfolio of defined AI ideas to take into the hackathon phase
  

The shift was practical. AI moved further into the fabric of daily operations rather than remaining a general productivity aid. Participants left with working assets and clear next steps.

What Happens Next

The strongest ideas from the innovation workshop will now progress into the GEFIE 3 Hackathon stage of the AI Transformation Playbook. There, small teams will iterate on concepts including:

• AI-assisted invoicing support

• Tender analysis tools

• Delay analysis and reporting support

• Chatbot use cases

• Logging systems for plant, equipment and engineer visibility
  

This structured progression ensures that experimentation continues with focus, ownership and measurable development.

In construction, small gains in coordination, documentation and analysis compound quickly across projects. The Value Discovery Day demonstrated that when AI literacy is in place, applied experimentation becomes both credible and commercially relevant.

Organisations that move beyond exploration and begin embedding AI within operational workflows position themselves for sustained improvement.

Explore how the GEFIE 2 AI Value Discovery Day supports applied AI transformation across operational teams: https://www.dixonai.com/transformation/gefie2-ai-value-discovery-day