The so-called “Vibe Coding” using LLM-driven tools such as Cursor Composer – a term coined by well-known computer scientist Andrej Karpathy – describes a handoff approach to writing code using the Genai model, and is really taking off these days. According to Y Combinator, a quarter of the latest batch startups rely on AI to generate 95% of the codebase. But despite its obvious speed, is it really the most efficient way to generate code? Can businesses leave unstable code and rising technical debt?
This is the view of UK-Startup Turintech, which launches Artemis products at Nvidia's GTC event. This is news along with news that it has raised a total of $20 million. Co-founded by Dr. Leslie Kanthan (CEO), Mike Basios (CTO), and Fan Wu (Chief Science Officer), Turintech describes Artemis as an “evolutionary AI” platform that can optimize and verify your enterprise codebase.
Founder Kanthan explained to TechCrunch: “Generating a lot of code generates a lot of inefficiencies. Use Artemis to find all of the inefficiencies in the generated code, faster and more resources.”
Unlike other Genai tools that rely solely on LLMS for basic code generation and optimization, Artemis aims to improve, validate and evolve your code to improve performance, security, and scalability.
Kanthan explained: “For example, Github Copilot predicts which code to write, but doesn't actually check the code, nor does it predict Cursor either. Traditional checking methods basically use a small tool that works with a compiler that checks the code for modifications.”
To support his claim, Qantan refers to his 2018 research paper. This caused some upset in the AI community by outlining Darwin's approach to data structure selection (and hence the term “evolutionary AI”). It is these ideas that make Artemis products built.
Turintech's latest funding round included an unreleased $15 million Series A, led by Oxford Capital, which was closed seven months ago and featured by Circle Rock and IQ Capital. Previously, they raised $5 million in seed funding.
While Kanthan was shy about naming clients publicly, TechCrunch provided evidence that Turintech already has a large blue-chip player and bank that took part in an early adoption program ahead of the full release of the platform later this year.
You might be pushing an open door. The code generated by AI is already transforming software development. By 2028, the Gartner project will use AI code assistants, up from under 10% in early 2023. Meanwhile, a stack overflow survey found that 63% of developers incorporate AI into their workflows.