Participated · Series B · Jan 2022

InstaDeep develops reinforcement-learning and advanced machine learning platforms to solve complex optimization problems for enterprises, from container routing to train scheduling and drug design. The company is headquartered in London with offices in Paris, Tunis, Lagos, Dubai and Cape Town and was founded in 2014. InstaDeep has over 170 employees, more than 130 of whom work in AI research, engineering, ML and DevOps, with roughly half the team based in African offices. Its product work includes a moonshot railway-scheduling project with Deutsche Bahn and an early warning system that detected high-risk SARS-CoV-2 variants ahead of WHO designations. InstaDeep has collaborated closely with BioNTech, Google (including DeepMind and Google Research) and others on applied AI projects. With the new funding the company plans to accelerate product launches across biotech, logistics, transportation and electronics, advance its computing infrastructure, expand into the U.S. and hire more talent. InstaDeep builds an AI-based enterprise decision-making platform that leverages deep reinforcement learning, machine learning, and predictive analytics to optimize decisions and improve ROI across manufacturing, logistics, mobility, and energy. The company says its platform can tackle challenging optimisation and automation problems in dynamic, complex environments and is already providing measurable value to clients. InstaDeep was founded in Tunisia in 2014 and is now London-based; it was one of the first African-founded AI firms to publish original research at NeurIPS in 2018. The startup has seen steady growth and has developed the platform through its own cutting-edge research. With its latest funding and a strategic partnership with AfricInvest, InstaDeep aims to expand advanced AI opportunities and talent development in Africa and beyond. Management highlights plans to democratise machine learning in Africa and leverage AfricInvest’s operational expertise to accelerate R&D and talent programs.