Through workshops with Sprint, KAFO and Replit, Fawaz Buqammaz is helping participants explore AI by building software, bringing the technology into the first stages of developing a business idea.
Fawaz Buqammaz’s work with aspiring builders in Kuwait starts with a practical question: what happens when someone with an idea can begin making an app before they have learned to code?
At the KAFO Buildathon, organised with Sprint and Replit, that question became a four-hour exercise focused on sports, nutrition and health. Sprint identified Buqammaz, Replit’s ambassador in Kuwait, as the session leader. Participants were asked to bring their laptops and work towards a prototype they could test.
The event, advertised for 26 September at Al Tijaria Tower, offers a concrete example of the kind of AI education Buqammaz is pursuing. Its immediate goal was a working first version. For people considering a digital business, that creates a way to examine an idea through something they can use, change and show to others.
A workshop built around making
Sprint’s account of the programme described a guided process in which participants used Replit to develop their ideas step by step. The three subject areas gave the exercise a defined scope, while the four-hour format set a clear limit on what participants were trying to produce.
It followed an earlier Replit Buildathon that Sprint’s Ghanim Aldosary described as Kuwait’s first. In his report from that event, Aldosary said participants arrived with ideas and left with working apps, with no coding experience required. He credited Buqammaz alongside Replit and its growth lead, Abdulrahman Aldhalaan, for helping deliver it.
These are organiser-reported outcomes. They establish that participants built software during the sessions, rather than showing how many subsequently launched businesses or attracted customers. That distinction helps explain the role of a buildathon: it gives an idea a first practical test, at a stage when the product itself is still being worked out.
Across the two events, Buqammaz’s role places him close to that early encounter between a prospective founder and the software they want to create.
Connecting technical knowledge with business ideas
Buqammaz approaches the subject through both computing and entrepreneurship. His published biography lists a bachelor’s degree in computer science with artificial intelligence from Northumbria University and a master’s covering business, innovation and sustainability from Newcastle University.
It also identifies him as co-founder and chief technology officer of QASD, where his work includes AI development and technical training, and as a co-founder of the educational initiative Ofoq. Those activities sit alongside his public explanations of AI and emerging technology.
The philosophy set out on his website is to make technical knowledge understandable without stripping away its substance. He emphasises explaining how systems work, where their limitations lie and how their outputs should be assessed. That is his stated approach, rather than an independently measured account of training results.
Read alongside the buildathons, it suggests a teaching method that connects explanation with an immediate task. A participant has to express what an application should do, inspect what appears and decide what needs changing. The learning therefore includes defining the problem, as well as operating the tool.
What AI changes about the first build
The technology behind these workshops helps explain their format. According to Replit’s documentation, its Agent accepts instructions in everyday language, writes code, sets up the project and tests the result. Users can then ask for changes through the same conversational interface.
This changes the starting task for a beginner. Instead of first assembling every technical component, the user begins by describing the intended application. Replit also provides a planning mode for discussing an approach before making changes, allowing the proposed work to be reviewed before construction begins.
Speed, however, needs to be assessed in context. In a February 2026 research update, the independent research organisation METR said it believed AI tools were accelerating experienced developers more than in its earlier study. It also said selection effects made the size of that improvement difficult to measure reliably.
Some developers were unwilling to participate when the experiment required them to complete work without AI. Others excluded tasks they particularly wanted to perform with AI assistance. Those choices complicated comparisons.
That research concerns experienced developers working on established software, rather than beginners at a Kuwait workshop. It does not measure Buqammaz’s sessions. It does show why producing a prototype quickly and demonstrating a general productivity gain are different claims.
What comes after the prototype
The next stage introduces responsibilities that remain with the person building the application. Replit’s shared responsibility model assigns users the task of checking generated code for correctness, security and licensing. The platform provides infrastructure and tools, but the application owner still makes decisions about its contents and configuration.
For a prototype, this means that a successful demonstration is one milestone in development. Before an application is used more widely, its creator needs to establish that it behaves as intended and handles information appropriately. A working screen alone cannot answer those questions.
For Kuwait’s emerging builders, the workshops offer an identifiable starting point: a defined problem, a tool and time to produce a first version. Buqammaz’s contribution is visible at that stage, where technical instruction meets an idea someone wants to try.
In these workshops, AI education takes the form of a build: define an idea, make a first version and examine the result. The open question is how many of those first versions progress into applications used beyond the workshop.