← NewsPortfolio news

Why we backed Disseqt AI

Disseqt AI has raised $8m to help enterprises deploy AI agents safely and at scale.

Every big shift in software has created a new layer that the rest depends on. The cloud needed security, and AI agents need assurance.

The biggest thing holding enterprise AI back today is trust, more than the models. Companies have plenty of pilots, but they can’t prove their AI will behave once it’s live. As agents move from answering questions to doing real work, inside the business and in front of customers, that proof is what everything else rests on.

We believe AI assurance will become a standard part of every enterprise’s AI stack, just as software testing and cybersecurity became for traditional software. It will sit inside how software is built and shipped, running all the time, not only once at the end. AI engineers, software engineers and researchers will use it every day as a normal part of their work. And it has to be independent, because you don’t let a company mark its own homework.

That’s the layer Disseqt is building, and the demand is already there. For a global sports league’s fan engagement agent, used by 80 million fans, Disseqt lifted the safe-response rate from 80% to 99%. It also provides assurance for a large European bank’s credit card programme. And it sells through global systems integrators and the platforms enterprises already use, so it meets customers where their AI is being built, rather than winning them one pilot at a time.

We first backed Apoorva, Manish and Cyril because they saw this need before the market did. Between them they’ve held senior roles at Microsoft, AWS and Salesforce, across enterprise software, AI and cybersecurity. Since we backed the company at pre-seed 12 months ago, sales have grown by 200%.

We’re delighted to follow on from our initial investment in the $8m seed round, led by Gresham House Ventures with Delta Partners, Enterprise Ireland and angels including senior leaders from OpenAI, Palo Alto Networks, Salesforce and Meta, and Rahul Dravid.

As agents take on more of the work inside every enterprise, the question will shift from whether to use them to whether they can be trusted. We believe Disseqt will be the answer to that question.