Case study
AI strategy, then an internal build that speeds up RFP responses
We worked with Knowit Cloud & Code's leadership to design an AI strategy and prioritize use cases, then led an internal development team to build the RFP matchmaker, which accelerates responses to IT project RFPs against a database of around a thousand consultant CVs and past projects.
- Matching time
- About a week, now a few hours
- Managers no longer polled per RFP
- Around 50
- Data scope
- Around 1,000 consultant CVs plus project history
- Built by
- Knowit's internal development team, led by us
- Preceded by
- AI playbook and prioritized use case portfolio
The situation
A consulting business answers RFPs for a living, and the quality of an answer depends on finding the right people and the most relevant past work fast. At Knowit Cloud & Code that meant searching across roughly a thousand consultant CVs and a history of delivered projects, under deadline, by people who could not possibly hold all of it in their heads. The wider question was where AI belonged across the business, for internal productivity and for what they offer clients.
What we did
Strategy first, then a deliberately chosen first build.
- AI strategy with leadership: an AI playbook covering both internal productivity and customer offerings.
- Use case discovery and prioritization, so the first build was chosen rather than defaulted to.
- Inspiration and enablement sessions for a wider audience, so the strategy was not held by five people.
- Leading an internal development team through building the RFP matchmaker: a tool that matches an incoming RFP against the consultant CV and project database, so the team responding starts from candidates rather than from a blank search.
Why the RFP matchmaker was the right first build
It sat squarely on the first path to AI value, boosting productivity in a workflow the business already understood and already measured. The data existed. The owner was obvious. The result was visible to the people doing the work within one RFP cycle. Choosing a first build with those four properties is most of what determines whether an AI program gets a second one.
What changed
A small team of salespeople now matches consultants to an incoming RFP in a few hours. The same work used to take about a week. The larger change is what stopped happening: answering an RFP no longer means sending a question to fifty team managers to find out who is available and who fits. The search runs against the database rather than against fifty people’s attention, which is the part that had a real cost and never appeared on any budget line.
What the client said
Amir worked with us to establish an AI Playbook for both internal productivity and customer offerings. He engaged with both leadership and technical teams, as well as held several inspiration and enablement sessions for a wider audience. Through a good combination strategic approach, bias for action, and technical understanding, Amir proved invaluable in accelerating our AI journey and achieving tangible results.
Johan RibberklintCEO, Knowit Cloud & CodeRelated: AI Use-Case Discovery and Prioritization · AI Implementation and Adoption