UK AI startup Cosine raises $2.5M for its AI developer outperforming human coders

UK-founded AI startup Cosine has raised $2.5 million from SOMA and Uphonest in a round that also included funds like Lakestar and Focal. In parallel, it achieved a breakthrough in AI-assisted software development with its artificial developer, Genie.

Based between San Francisco and London, Cosine's Artificial Developer, Genie, works like a very good human developer. 

For instance, it is able to solve bugs, build features, refactor code, and everything in between either fully autonomously or collaboratively with other developers. 

Founded in 2022, Y-Combinator-backed Cosine's software was created out of the founder's realisation of the potential in using LLMs to perform complex tasks in the coding space by imitating human software developers' behaviours. It is uncannily 'human' in its approach to reasoning as a result, with the founders' primary goal to create truly resilient AI capable of tackling open-ended problems across various domains

The company has achieved a 30 per cent score on SWE-Bench, the industry standard for evaluating software engineering skills in AI models. 

The benchmark, which includes real-world human tasks in software architecture, debugging and implementing new features in existing codebases, assesses an AI model's ability to understand, modify, and generate complex code. This marks the highest score achieved by any company to date 

Cosine's score represents a 56 per cent improvement over the previous best score, held by Factory at 19 per cent and 2196 per cent improvement over OpenAI's GPT4 score of 1.31 per cent. 

By fine-tuning models to emulate human reasoning, Cosine's approach has beaten out rivals like AWS's Amazon Q Developer and Cognition's Devin, both of which scored under 20 per cent  on the same benchmark.

Cosine CEO, Alistair Pullen, shared: 

"Our breakthrough in codifying human reasoning is allowing us to train AI models to operate far beyond the narrow range of tasks and tightly restricted prompts currently available to teams developing software." 

"We've developed a product capable of beating OpenAI and others in completing complex software tasks - in a fraction of the time and money it has taken our competitors to achieve the same results. We're on course to radically transform the way development and developers work", continued COO Yang Li. 

"Cosine is not just improving AI; they're fundamentally teaching AI to reason, providing companies with a true AI colleague", said Ellen Ma, Partner at Uphonest Capital.