The problem this solves
Most organizations already have AI ideas. The problem is that a typical strategy session generates 30–80 candidates with no framework for deciding which ones to build. The list ends up in a slide deck. Six months later, the organization is still discussing the same ideas, with a few new ones added to the pile.
The bottleneck is rarely ambition. It is the absence of a decision mechanism. Without a scoring framework, prioritization defaults to whoever speaks loudest in the room. Executive pet projects advance. High-value, high-feasibility opportunities that don’t have a vocal champion get buried. The organization spends budget on AI initiatives that were never the right starting point.
The Discovery Sprint replaces that cycle with a structured, facilitated decision process. In a single half-day session, your leadership team moves from a long list of AI possibilities to a scored, ranked shortlist of 2–3 use cases that are ready for architecture design, and assessed for AWS funding eligibility on the same day.
Who this is for
You have a mandate to deploy AI but no consensus on where to start. You need a defensible decision, not another ideation workshop.
You've collected use case ideas from across the business. You need a structured framework to prioritize and eliminate the noise.
You're responsible for the processes AI would affect. You need to be in the room when the prioritization decisions are made.
How it runs
The scoring framework
WSJF: Weighted Shortest Job First, is a prioritization method developed in scaled agile delivery and adopted widely in enterprise portfolio management. It ranks work items by dividing the cost of delay by the effort required to deliver. In plain terms: it surfaces the ideas where waiting is expensive and building is achievable.
In the Discovery Sprint, WSJF is applied to AI use cases across four dimensions: business value, time criticality, data readiness, and implementation effort. Each use case receives a numeric score. The scoring is done collaboratively in the room, with every participant working from the same criteria. The result is a ranking that reflects business reality rather than organizational politics.
Data readiness acts as a hard filter before final ranking. A use case with a high WSJF score that depends on data that doesn’t exist, isn’t accessible, or has quality problems is deprioritized. This prevents the most common failure mode in enterprise AI: committing to a build before the data foundation is confirmed.
What the output looks like
The Tier 1 shortlist is the direct input to the Concept Sprint. Each item in Tier 1 has a clear hypothesis, an identified data source, a named business owner, and a funding assessment. It is not a list of suggestions, it is a decision record.
What “done” looks like
By end of session, you have a signed-off shortlist, not a "next steps" slide. Every item includes a hypothesis, a data source, an estimated build window, and a funding pathway.
2–3 Tier 1 use cases ranked by WSJF score, each with a named business owner, a defined success metric, and a confirmed data source.
A structured view of your organization's current data infrastructure, tooling, and team capability as it relates to the shortlisted use cases.
A clear assessment of which shortlisted use cases qualify for AWS PoC funding (up to €10k) or AWS Migration Acceleration funding (up to €400k).
A complete record of the session: all ideas generated, scoring rationale, decisions made, and the criteria used, shareable with stakeholders who were not in the room.
The Discovery Sprint replaces the workshop nobody acts on with a session that produces a decision the same afternoon.
Frequently asked questions
"We'd been discussing the same AI ideas for six months. The Discovery Sprint gave us a scored, signed-off shortlist in half a day. The funding assessment alone saved weeks of procurement back-and-forth."
What makes this different from a strategy session
A strategy session produces a presentation. The Discovery Sprint produces a decision. That distinction matters because a presentation requires another meeting to act on it. A decision does not.
The difference is methodology. Every idea generated in the session is immediately scored against four criteria, filtered through a data readiness check, and assessed for funding eligibility, before the session closes. There is no post-processing, no follow-up workshop to interpret the outputs. The shortlist is signed off in the room, on the day, by the people who will own the work.
This is also not a discovery exercise run by generalists. Linda Mohamed has 12+ years in enterprise IT and holds AWS Community Hero status, one of fewer than 300 globally. Every session is grounded in what AWS actually funds, what enterprise data architectures can realistically support, and what the gap between ambition and execution typically looks like in practice. The scoring reflects that experience, not just a methodology pulled from a playbook.
Four deliverables. One half-day session. The day starts with a long list and ends with a defensible decision.