Evoid is a small venture I co-founded to do applied AI and computer-vision work for clients in Beirut, alongside a team that includes Tayseer Laz and Abou Baker Al Khatib. I worked as systems engineer and product manager: client coordination, technical direction, and building the public site.
The useful lesson was about scope. A client with a workflow that a camera could improve rarely needs a platform; they need one bounded prototype that answers one question, delivered by the same people who will maintain it. PadelEye, the venture’s own experiment in automated padel line judging from a camera stream, is a case in point: trained weights and a scaffolded pipeline exist, and it is still a prototype, not a product.
Between 2023 and 2025 we delivered five computer-vision applications for external clients. I don’t name them or their outcomes here; that is for each client to agree to first.