AI strategy for an accessibility app.
An accessibility technology platform · founder-facing AI advisory
The founder knew the product and had a fixed budget to build it with, without the engineering background to buy it safely. OpenKit advised on the AI side of it, from the first prototype through to choosing the developer who would build the platform.
Overview
Wakari is an accessibility technology platform. The idea was to replace a spreadsheet-based questionnaire with something people could answer in their own words, and the founder had the product clarity for it without an engineering background. The budget was fixed and had to cover the whole build, which put weight on every early decision.
OpenKit's part was strategic AI consulting and an implementation roadmap, shaped around keeping the founder in control of what the budget was spent on.
- The idea earned its place on throwaway prototypes, before any development budget went out.
- The founder came out of it able to read a quote and challenge a recommendation.
Advisory only
Strategic AI consulting and an implementation roadmap for the founder
Fixed budget held
The work stayed inside the original fixed budget
Platform live
Wakari launched after OpenKit's work finished, built by the developer the founder chose
Approach
How the advisory ran
OpenKit worked through the decisions in the order the founder had to make them, starting with whether the idea held up at all.
Prototypes that tested the idea cheaply
A custom GPT and low-code prototypes put the spreadsheet-to-conversation idea in front of real people.
AI vocabulary for a non-technical founder
OpenKit went through how language models work, where they fail and what they cost to run, which is the vocabulary a founder needs to brief a developer without an engineer translating.
A framework for choosing a build partner
OpenKit wrote a red-flag framework for judging development partners, so the founder could test a pitch against evidence.
A staged route the budget could carry
Scope documents and a staged validation route the founder could hand to a developer, with each stage proved before the next one was paid for.
Controls
What the advisory used
- Custom GPT prototype path
- Low-code prototyping
- Proof-of-concept evaluation
- Vendor-management framework
OpenKit certifications
- ISO 27001
- ISO 9001, UKAS-accredited
- Cyber Essentials
Controls on this project
- WCAG 2.1 AA target
Other engagements
What could your business do with AI?
Find out with experts who become part of your team. We uncover opportunities, get ideas working, and help your people build on the results. We reply within one working day.
Get in touch AI Audit