Cambridge-based startup Hiverge has emerged from stealth with $5 million in seed funding to develop what it describes as an “algorithm factory” that uses AI to automatically generate and optimise algorithms for businesses. The company was founded by former Google DeepMind researchers who previously worked on breakthrough projects including the Nobel Prize-winning AlphaFold protein prediction system.
The seed funding round was led by Seattle-based Flying Fish Ventures, with participation from Ahren Innovation Capital and Alpha Intelligence Capital. Notably, Google’s chief scientist Jeff Dean also invested in the startup, lending significant credibility to the venture.
Founding Team with Distinguished Pedigree
Hiverge’s founding team brings extensive experience from Google DeepMind’s most prestigious projects. CEO Alhussein Fawzi is a former Google DeepMind research scientist and MIT Innovators Under 35 award winner who led the team that developed AlphaTensor, the AI system that discovered new matrix multiplication algorithms in 2022.
Chief Technology Officer Bernardino Romera-Paredes was part of the Nobel Prize-winning AlphaFold team that revolutionised protein structure prediction. The founding team also includes Chief Science Officer Hamza Fawzi, a professor in the Department of Applied Mathematics and Theoretical Physics at the University of Cambridge, specialising in mathematical optimisation.
Addressing Critical Business Algorithm Challenges
The company aims to solve what it identifies as a hidden yet critical business challenge: the discovery and deployment of algorithms that help organisations rapidly improve their operations at scale. Current algorithms often fall short of today’s requirements as systems grow in complexity and businesses need to make accurate decisions in milliseconds whilst computational resources remain constrained.
Alhussein Fawzi, CEO of Hiverge, explained the market opportunity:
“Algorithms drive the revenue of most businesses today – from finance to AI inference. But algorithms of today are struggling to keep up with challenges of scale and speed, and traditional methods aren’t working anymore. At Hiverge, we are creating algorithmic superintelligence, giving companies the ability to discover and deploy not just new algorithms, but ones that are tailored to their specific tasks. With Hiverge, companies will be able to unlock algorithms that are beyond the limits of current approaches.”
The Hive Platform and Early Results
Hiverge’s platform, called the Hive, uses program synthesis to create what the company terms an “algorithm factory” – automatically writing and optimising code tailored specifically to each user’s unique challenge at scale. Unlike traditional software engineering approaches, the technology combines AI creativity with rigorous testing to discover breakthrough algorithms.
The platform has already demonstrated impressive results across multiple domains. The Hive discovered fast and resource-efficient training algorithms that improved over existing human-engineered approaches by more than 15% on established AI benchmarks. Additionally, on an industrial planning problem posed by Airbus, the platform discovered sophisticated long-term planning algorithms that run several orders of magnitude faster than existing classical planners.
The company successfully demonstrated its capabilities by winning the Airbus Beluga Challenge, developing a solution that delivered a 10,000-times speed-up over the existing aircraft-loading algorithm.
Industry Validation and Partnerships
Hiverge has already secured validation from industry leaders, including quantum computing company Quantinuum. Konstantinos Meichanetzidis, Head of Product Development and AI Strategy at Quantinuum, highlighted the platform’s potential:
“Our quantum research demands highly specialised algorithmic approaches. We’re excited to be collaborating with Hiverge and using its platform to discover novel quantum algorithms. We see strong potential for this partnership to accelerate the scientific breakthroughs we’re pursuing.”
Investor Confidence and Market Positioning
The funding round attracted significant investor interest, with Flying Fish Ventures leading the investment. Frank Chang, Managing Partner at Flying Fish Ventures, expressed strong confidence in the company’s vision:
“It usually takes us some time to get conviction to invest, yet with the Hiverge team we knew immediately we needed to back the company’s vision. Hiverge is delivering everything we look for in a company: a highly experienced founding team with deep technical expertise in their field; a genuinely innovative solution to a major market issue; and a unique deployment of powerful technology to enable it all. This is a completely new category that will redefine how algorithms are deployed by businesses.”
Technology Heritage and Innovation
The Hiverge team’s work builds on pioneering research at Google DeepMind, where they contributed to projects including AlphaTensor, FunSearch, and the early development of AlphaEvolve. AlphaEvolve notably boosted the efficiency of Google’s data centres, chip design, and AI training processes, with Google crediting the system with finding ways to optimise its large language models, including a 23% speed-up for Gemini’s matrix multiplication and a 32% boost to Transformer operations.
Commercial Strategy and Future Plans
The funding will be used to accelerate go-to-market activities, expand research capabilities, and support product development. The company plans to make its technology accessible through cloud marketplaces like AWS and Google Cloud, where customers can directly use the system on their own code.
Hiverge currently employs six people but plans to expand its team whilst transitioning from research to building out its commercial product. The platform will analyse code bottlenecks, generate improved algorithms, and provide recommendations to engineers, democratising access to advanced algorithmic capabilities that have traditionally been available only to large companies with extensive data science teams.
Businesses interested in accessing the platform can join Hiverge’s early-access waitlist as the company prepares for broader commercial deployment of its algorithm factory technology.
