Cogna Raises $15M for AI-Powered ERP Software Development

From self-driving cars to AI that writes enterprise software: Cogna founder raises $15M

A founder who was an early mover in the race to build autonomous vehicles has raised $15 million for his next act: a startup that claims its AI can write enterprise software on its own. Cogna -- as the U.K.-based startup is called -- is led by Ben Peters, the technical co-founder of FiveAI (the self-driving startup that was acquired by Bosch in 2022). Notion Capital is leading the Series A, with Hoxton Ventures and Chalfen Ventures also participating. It comes on the heels of Cogna, founded in May 2023, raising a seed round of $4.75 million earlier this year from a number of investors, including Peters' FiveAI co-founder Stan Boland and Herman Hauser, the founder of Acorn Software. The focus of Cogna is the world of enterprise resource planning. ERP is a dry but very necessary software component in the running of organizations, covering everything from procurement and supply chain and inventory management through to risk assessments, finance and human resources. Typically large enterprises can pay up to billions of dollars in contracts with systems integrators and consultancies to handle their ERP, either by customizing off-the-shelf software for clients, or writing custom applications from the ground up, to fit an organization's particular needs. In typical AI startup fashion, Peters believes that this work is best done, and can be compressed into, an AI platform built for the purpose. While some challenges have been significantly trickier for AI to address than others, there have been some early signs of ERP potentially being one area where it might stick, and for Cogna to potentially become one of the players delivering on that pitch. A year since launch, Cogna has signed up customers that include the U.K. gas distributor Cadent Gas and infrastructure and utilities service provider Network Plus. "To be clear, none of our customers think of it as ERP software," Peters said in an interview. "For them, they've got a problem which isn't solved by using SAP or [another] legacy system. We deliver a custom, precision-built piece of software for that. They experience [it] ike a classic piece of SaaS, but it happens to have been built specifically for them, specifically for their workflow. And it's written by our AI." Cogna balances a mix of generative AI with other kinds of tooling to actually put together its custom software. Non-technical teams can describe their pain points in natural language, and the Cogna does the rest, the company claims. "We're a team of experts in domain-specific languages, compilers, AI and reliable and scaleable enterprise SaaS," Peters said. These days, it's not uncommon to come across AI startups leveraging multiple large language models simultaneously depending on a particular task at hand -- it's the customization that makes a startup unique -- and that is the case here, too. Peters said that the language models used for people to interact with its platform come from a variety of providers that include OpenAI and Anthropic. These are "a key part of our 'Natural Language Compiler," Peters told TechCrunch over a call. On top of that, Cogna is building an engine so that "the software can truly write itself," said Hussein Kanji, the co-founder of Hoxton Ventures. In doing so, Cogna is an example of how generative AI -- which saw a viral explosion with the launch of consumer-accessible services like ChatGPT -- is indeed being folded into enterprise usage and more complex applications. That is important because a number of foundational AI companies, like OpenAI, have made it clear that they, too, have ambitions to tap into the enterprise market, a highly lucrative area that will help them generate the vast revenues that investors expect. Kanji likens what Cogna is doing to Cursor, the code building tool from OpenAI-backed Anysphere. We broke the news last week that Anysphere is in the middle a massive, unsolicited bidding war among investors that want to back it, valuing Anysphere as high as $2.5 billion on the back of massive business growth, so the area is clearly very hot. "That's kind of what Cogna is also doing," he said. Based on his' experience at self-driving tech company FiveAI, Peters is adept at spotting openings in the market early, but he also has experience in how to shift gears, so to speak, when plans don't go as you expect them to. "We had cars driving around London in 2018," he recalled. "[But] we then pivoted to providing the development environment to develop a stack for other companies that were building their own electronic design automation stack. We were running a million simulations a day by the time we sold the company." Five.ai, which had raised around $78 million in funding was valued at $216 million in its last round before it was acquired by Bosch for an undisclosed sum. Cogna now plans to go up against classic IT consultancies such as Wipro and CapGemini. The investment into Cogna is the first from Bryan Gartner, formerly of Khosla Ventures, who joined Hoxton as its newest partner last year.

Cogna reels in $15M for its AI-powered ERP platform - SiliconANGLE

Cogna Ltd., a UK-based enterprise resource planning startup, today announced that it has raised $15 million in fresh funding. Notion Capital led the Series A round with participation from Hoxton Ventures and Chalfen Ventures. The raise comes a few months after Cogna closed its initial $4.75 million seed investment. The latter round included the participation of Hermann Hauser, a prominent tech entrepreneur who played a key role in launching Arm Holdings plc. Enterprise resource planning, or ERP, platforms are complex software suites that companies use to run their day-to-day business operations. Such products' features set usually cover sales, marketing, procurement, regulatory compliance and a wide range of other use cases. Companies buy ERP software partly because it eases data management tasks such as sharing business information between departments. The typical ERP platform can take upwards of months to deploy. Cogna, in contrast, is promising setup times as short as two weeks. This speedup partly stems from the fact that the company uses artificial intelligence to speed up certain aspects of the configuration process. Cogna's platform ships with a chatbot-like interface powered by large language models. Workers can enter natural language descriptions of the business tasks they wish to perform more efficiently. Using the provided information, Cogna's platform generates a custom software tool capable of streaming the specified chores. A customer could, for example, inform the platform that the accounting team is struggling to find the best way of splitting a budget among three application projects. In response, Cogna might generate an analytics tool that makes it possible to simulate different budget allocation scenarios. Workers can optionally use natural language instructions to customize the tools that the software generates. Business planning is not the only task that Cogna promises to ease with its platform. A company's external documents, such as contract bids submitted by suppliers, are often formatted in different ways. That complicates tasks such as identifying data entry errors. Using Cogna, a company can create a portal that allows suppliers and other third parties to submit documents in a uniform format. Field service is another use case that the software maker is targeting. According to Cogna, organizations can leverage its platform to share data about a piece of malfunctioning equipment with the technicians sent to repair it. The company's ERP platform also lends itself to other use cases spanning human resources, regulatory compliance, customer support and several other fields. Many of Cogna's customers operate in asset-heavy sectors such as the utility, manufacturing and construction segments. The company says that its software has helped some of those organizations realize more than $10 million in annual cost savings. Cogna is not the only ERP startup to have raised venture funding recently. In August, Opkey nabbed a $47 million Series A round for its namesake software testing application. The tool helps companies check updates to their ERP deployments for errors before rolling out them out to production.