Language
Computing power has become the biggest barrier to meaningful AI progress in the MENA region. By participating in this program we're unlocking the infrastructure needed to accelerate real innovation and position ourselves to serve 250-400 new enterprise clients in the next two years.
WideBot AI is building the Arabic-first AI agents platform powered by its own 7B LLM AQL. It delivers enterprise-grade voice and language technologies that automate real operations from customer experience to banking healthcare and telecom use cases. WideBot's agents are already deployed across multiple government entities and leading enterprises in Saudi Arabia and the wider region proving scalability and real business impact. Founded by Mohamed Nabil & Mohamed Mostafa and a technical team experienced in large-scale model training using PyTorch and vLLM and multi-cloud deployment across AWS Google Cloud and Runpod the company addresses a critical regional gap: limited Arabic dialect coverage and enterprise-grade retrieval over Arabic knowledge bases. By combining advanced text-to-speech (TTS) speech-to-text (STT) and retrieval-augmented generation (RAG) models trained on local Arabic datasets WideBot AI helps enterprises and public sector teams deliver natural Arabic voices accurate transcription and secure domain-aware responses. Through the Compute Accelerator Ready Track the team will optimize models using dedicated GPU and cloud credits from Cineca AWS and Google Cloud. With growing demand across Saudi Arabia the UAE and Egypt WideBot AI aims to serve the wider Middle East by March 2026 turning access to compute into real-world impact.
