AI-powered accessibility. Three funding rounds. Two prior exits. One very specific problem to solve: the internet still doesn't work for 1.3 billion people.
Here's a number that should bother you: Fortune 500 company websites average nearly 20 accessibility errors per homepage. And homepages are the easy part - the pages that get the most scrutiny, the most design review, the most QA. What happens deeper inside, behind login screens, inside transactional flows and dashboards? Navin Thadani spent years figuring out the answer. He didn't like what he found.
Thadani launched Evinced in August 2018 in Palo Alto alongside co-founder and CTO Gal Moav, with a precise diagnosis: the accessibility problem wasn't a design problem or a legal problem. It was a developer workflow problem. Accessibility testing was happening too late - after code shipped, after design was finalized, after lawyers got involved. Prevention, not remediation, was the only path that actually scaled.
That framing - developer-first, prevention-first, automation-first - turned out to be exactly what enterprise buyers had been waiting for. By the time Evinced closed its $55M Series C in December 2024, the company was serving five of the ten largest media companies in the US and UK, three of the five largest B2B SaaS companies, three of the ten largest healthcare companies, and financial institutions collectively managing $26 trillion in assets.
"Homepages are often the easiest pages to make accessible because they are static and heavily reviewed. The bigger challenge and the bigger opportunity lies deeper inside authenticated workflows, applications, dashboards, and transactional experiences where accessibility problems become harder to detect and more impactful for users."
What Evinced built is not a compliance checklist tool. The platform uses computer vision, machine learning, and generative AI to detect, cluster, and classify accessibility issues directly inside development pipelines - integrating with the tools developers already use. Playwright, Appium, Espresso, Jira, Slack. The product catches problems before they ship, not after the lawsuit arrives.
Thadani calls it the only "pure tech play" in the accessibility space. He means it literally. Most competitors are services businesses wrapped in light software. Evinced is engineering-first, with what Thadani believes is the largest accessibility engineering team in the industry.
Thadani came to Silicon Valley via Pune and Stanford. A BS in electrical engineering from the University of Pune gave him the technical foundation. An MS in economics and operations research from Stanford gave him the systems-level lens. An MBA from Wharton gave him the business grammar. Three disciplines, three institutions, and more than 20 years of building things that get acquired by companies with names like Red Hat and Oracle.
His first founding-team role was at Qumranet, a virtualization startup where he served as GM of the Virtualization Business Unit. Red Hat acquired Qumranet in 2008. He joined Ravello Systems next, another founding team, this time as VP of Product Development. Oracle acquired Ravello in 2016 for a reported ~$500 million. He then spent time in product leadership at Oracle post-acquisition - the kind of experience that either sharpens a founder's appetite or kills it. For Thadani, it sharpened it.
Two years after the Oracle chapter, he co-founded Evinced. The accessibility market - then dominated by manual auditing services and basic rule-based scanners - was ripe for exactly the kind of technology-first disruption he had learned to execute.
Evinced's early technical bet was that computer vision could detect accessibility issues that rule-based tools missed - visual problems invisible to HTML parsers but obvious to a user with low vision trying to navigate a complex UI. That bet, made in 2018, looks prescient now. The company has since expanded its AI stack to include generative AI for root cause analysis, issue clustering, and automated fix suggestions.
In May 2026, on Global Accessibility Awareness Day, Evinced launched the Evinced 500 - a benchmark analysis of Fortune 500 company websites that exposed, with data, just how far even the most resourced organizations have to go. The finding wasn't surprising to Thadani. The specificity of the data was the point: if you want enterprises to move, you show them exactly where they stand relative to peers.
Evinced's research also turned up a quieter problem: leading large language models have accessibility blind spots. The AI tools being used to build the next generation of the web are themselves failing users with disabilities. It's a recursive problem that only a company thinking about accessibility at the infrastructure level is positioned to solve.
Insight Partners led Evinced's December 2024 Series C, with participation from M12 (Microsoft's venture fund), BGV, Capital One Ventures, Engineering Capital, and Vertex Ventures Israel. The round brought total funding to $112 million. The capital is funding European expansion, deeper generative AI integration, and growth in global sales and customer success teams.
The company has offices across the US and in Tel Aviv, where co-founder and CTO Gal Moav runs R&D. Thadani manages the overall enterprise from Palo Alto, where the pitch to Fortune 500 companies has evolved from "you should care about accessibility" to "here is exactly how many issues you have, where they are, and how your AI pipeline is making them worse."
That shift - from moral argument to technical specificity - is the Thadani approach in miniature: find the precise problem, build the precise tool, show the precise data. It's the same instinct that drove him from Pune to Stanford to Wharton to two exits and now to one of the most funded pure-play accessibility companies ever built.
"We are never going to be able to get to a good place in accessibility unless we focus on this notion of prevention."
We are never going to be able to get to a good place in accessibility unless we focus on this notion of prevention.
Accessibility is too important to be put off so late in the cycle. It needs to be as much a part of a developer's job as making sure their code works.
Nobody is really looking at it the way we are; we are still, at this point, the only pure tech play in the accessibility space.
In technology, it's often the case that it's hard to make something easy, and these funds are going to enable us to expand our engineering commitment, already we think the largest in the industry, to levels the industry hasn't even thought of, much less seen.
Homepages are often the easiest pages to make accessible. The bigger challenge lies deeper inside authenticated workflows, applications, and transactional experiences where accessibility problems become harder to detect and more impactful for users.