Ahmed Medhat

Co-founder

2025 - Current

Solving Collaborative AI. I think the most critical problem to work on today is to steer AI towards acting as an expanding force of intelligence abundance, not a deflationary force of AI replacement of human labor.

Computational Neuroscience - NeuroAI

2023 - 2024

Built a BERT based transformer model pre trained on single cell transcriptomes from the mouse brain’s neocortex. Model was used to predict electrical properties and locations of brain cells, strictly based on their genetic signature.

Angel Investor

2016 - 2024

Angel Investor, shareholder & advisor in 25+ startups since 2016. Commonly providing the first cheque for a startup. My investment thesis is supporting startups that solve operational and consumption inefficiencies and problems grounded in lack of access to information. This has led me to make diverse investments, including measurement-centered recruitment platform (Hackajob), health sensors (Cloudcath), robotics for lab automation (Automata, first cheque) and smart meter integration (Pylon). I do a combination of investing personal money, leading syndicate deals and acting as a venture partner to early stage funds. I also occasionally act as a hands-on advisor, helping with fundraising growth rounds, building data products and teams, and in capturing the necessary network effects for growth. As of May 2025, my investment portfolio has achieved an IRR of 43.1%.

Staff Research Scientist, Graph Learning

2020 - 2023

Built graph neural networks and network clustering algorithms for modeling user behavior and preferences on an Instagram graph of a size exceeding 100s of billions of connections. As part of a task force aiding in Covid response, built an epidemiological model of COVID spread, by modeling global patterns of collocation across different regions of the world. Our work was used by WHO and hundreds of other organizations as part of responding to the COVID pandemic, and was published in the journal Epidemics.

Principal Data Scientist

2015 - 2020

As a lead on ecosystem data science, I took on a role that was part ML, part causal inference, part network science and part growth hacking. Through my five years in that role I became the data lead and subject matter expert in Facebook on original content sharing. In terms of both what drives it and how to grow it. I led projects that investigated drivers of content sharing behavior to motivate creation of products that fulfill user’s sharing needs. • 2017-2020 key project: Ran a 3 year effort to create ground-truth data and classifiers for understanding content creation motivations on Facebook. I conceived of the project, secured funding for it, and managed 100s of labelers over the span of two years. • 2015-2018 key project: Investigated what drives people to share less or more content on Facebook. This work drew on on graph learning, mass communication theory, causal inference and network experimentation, to quantify how audience size, perception biases, novelty effects and competition contribute to a person’s decision to share content. Such as how the friendship paradox shapes a person’s sharing rates due to perceiving their friends receiving more feedback than they actually do. This both influenced how newsfeed posts are designed and led to two peer reviewed papers bringing an understanding of how the network structure between people can decide whether collectively they will share more or less original content.

Founding Chief Data Officer

2011 - 2015

As the founding Chief Data Officer, my role revolved around building the company’s data science capabilities to uncover private company data and understand user behavior. I helped steer Duedil's growth to one of the leading Fin-tech Startups in Europe, growing the team by 10X, users by 50X and launching our paid offering in the process. Things I did; - Led the team that built DueDil’s core in-house proprietary datasets including company-website matches, company’s trading names, keywords and industry classifications, by collecting, processing and normalizing data gathered from third parties, the web and public records. - Designed an end-to-end analytics infrastructure involving daily collection of data generated within the company and from user activity, and serving that data for interactive querying by analysts, dashboard users and engineers across the business. - Attracted top data science, analytics and credit risk talent from leading universities like Cambridge, Oxford and University of Illinois at Urbana Champaign, and companies like Facebook and Morgan Stanley. - As a member of the executive team, I played a key role in; o Creating series A and series B fund raising materials, and pitching to and securing investments from Tier 1 VCs in the US and Europe. Helping raise a total of $22 million. o Devising our business plan and go to market strategy. o Securing key partnerships with third party providers of proprietary, untapped sources of transactional private company information. o Leading various high-profile research projects related to issues of public interest, including an influential research piece on Migrant Enterpreneurs and an analysis of UK digital businesses sponsored by the Cabinet Office.

Researcher

2011 - 2013

Software Engineering Intern

2008 - 2008