From Ideas to Funded AI Work in One Accelerator Lab

A proven, AI supported process to turn Essential Energy’s ideas into a shortlist of costed, ready to move initiatives

“I’ve never been in a workshop where so much pre-work translated so quickly into strategic initiatives. The agents completely changed the pace.”
— Song Yeap, AI Lead - Barwon Water

Building a single, governed funnel for your AI portfolio

A structured, accelerator pattern to organise scattered AI activity into a coherent portfolio, with common language, clear priorities and guardrails that fit Essential Energy’s AI strategy

Two kinds of needs we see at Essential Energy
1. Strategic need

Clarity and confidence

Leaders need to see which AI opportunities genuinely move the dial on safety, reliability, customer experience and efficiency. You want fewer slide packs and more investable options with transparent sizing, risks and trade offs.

2. Execution need

Momentum without control

Organisations need help turning good ideas into delivery ready packages. They are looking for guidance and structure to ensure they can get started and invest in the right AI initiatives.

The acceleration workshop is designed to bridge these two needs. It is an AI assisted event that uses AI assistants to generate self service pre work at scale, giving Essential Energy a repeatable way to progress AI work.

This video is a short-form edit of a livestream we ran to showcase a live success story at Barwon Water, a fast-growing utility, facing demand it can’t meet by simply adding more people, used an AI-supported accelerator day to decide which AI initiatives to back.

In the video, we cover

  • Set in the context of a client engagement with Barwon Water, a fast-growing water utility.

  • Two hard truths: idea overload with no prioritisation, and scattered pilots that don’t add up.

  • We demonstrate two AI accelerator agents in action: one for structured intake, one for rapid sizing.

  • A single funnel channels ideas and pilots into a few strategy-aligned, delivery-ready initiatives.

  • 100+ raw use cases are turned into structured, T-shirt-sized options leaders can compare and back in the room.

  • The day compresses months of BA and portfolio work into a clear, governed AI roadmap leaders can act on immediately.

For the full livestream, including extended Q&A and deeper dives into the tools, facilitation and Barwon Water story, you can watch the full session here

The acceleration workshop playbook

How the accelerator sprint works

In one focused sprint you move from a long, messy list of ideas to a tight portfolio of costed, risk framed initiatives, agreed owners and a clear runway for delivery. Leaders see where to place the next bets. Teams leave with momentum, shared language and artefacts they can reuse as they integrate and scale priority AI use cases.

0

Pre work: capture and clean ideas

  • Participants receive briefings on the day and an overview of the self service pre work.
  • An AI Business Opportunity Assistant structures each idea, covering problem, strategic alignment, value, owners and implementation milestones.
  • You start the day with rich, comparable ideas, not a blank whiteboard.
1

Welcome and shared frame

  • Reset on strategy, constraints and success criteria.
  • Agree the rules of the game and how decisions will be made.
  • Clarify what happens to the AI business opportunities after the workshop.
2

In room review and first stage of funnel

  • Execs, business SMEs and AI enthusiasts work together to review and refine the pre captured ideas for their business area.
  • Silent digital voting narrows many ideas into a shortlist per table.
  • Everyone can see how and why ideas move forward or drop out.
3

Rapid sizing with the solution architecture assistant

  • Shortlisted opportunities are broken into components using an AI solution architecture assistant.
  • Effort estimates are applied to each component using a consistent, rules based model.
  • Solution components are rolled up to consistent T shirt sizes, allowing direct comparability across all the business opportunities.
  • The assistant also proposes a low complexity way to get started, using simple self service tools (such as ChatGPT) and reduced functionality.
4

Portfolio view and trade offs

  • Leaders and teams see the set on one page, with value, risk, cost and time.
  • Explore what if portfolios such as low risk starters, bold bets or balanced mixes.
5

Playback and decision moment

  • AI assisted coding rapidly creates solution mock ups, supporting short, structured pitches for the strongest candidates.
  • Sponsors give live direction to progress, park or pivot.
  • The room aligns on a small number of lighthouse initiatives to move first.
6

Retro, synthesis and next steps

  • Fast retro on what worked and what to change.
  • We play back distilled insights, patterns in ideas, risks and enablers.
  • Next steps are locked in, including owners, support model, milestones and measures of impact.
The pattern is repeatable. You can rerun the sprint as your AI pipeline grows, keeping strategy, architecture and teams aligned on what moves next.

Accelerator agents

To make the day faster and more concrete, we use two custom accelerator assistants.

Business Opportunity Assistant

An AI assistant performing the role of an experienced business analyst, helping to ask the right questions and structure the answers to rapidly complete a business requirements document with high quality content. It asks a structured flow of questions on the opportunity, strategic alignment, value, owners and implementation milestones. A process that traditionally takes weeks, with multiple meetings and document revisions, is completed in minutes.

Product Requirements Assistant

An AI assistant to help break down the idea and assign a T shirt size. It ingests the output from the Business Opportunity Assistant, clarifies terminology and then uses AI to propose a complete list of functional components required to deliver the idea, such as integrations, data requirements and application screens. The output is a concise product requirements document with T shirt sizing and an MVP suggestion. The document allows ideas across the AI portfolio to be compared for size, complexity and ability to get started, helping accelerate prioritisation decisions from weeks to hours.

Business outcomes

  1. Sharper investment decisions – A transparent, side by side view of options, value, risk, cost and time, so leadership can choose with confidence.

  2. Faster path from idea to action – AI supported sizing compresses weeks of architecture and costing into hours, without skipping the necessary checks.

  3. Stronger alignment across functions – Operations, digital, finance and customer teams work to the same facts and criteria, reducing friction later.

  4. Reusable pattern, not a one off – The sprint design, prompts and artefacts can be reused for future themes, building an internal “AI portfolio engine.”

  5. Capability built as you go – People learn how to frame, stress test and progress AI use cases in the context of their real work, not a classroom exercise.

We are looking forward to reconnecting to tailor this event to Essential Energy’s specific requirements and challenges.

Jack Steele
Exec Director - AI and Data, FromHereOn

Let's Talk