He builds digital identities out of the open web - and sells them to the banks that thought they already knew who you were.
A credit bureau looks at a paper trail. Idan Bar-Dov looks at a digital one - the aliases, the reused phone numbers, the reputational exposure scattered across the public web - and asks the question the old files cannot answer: is this person who they say they are, right now, this second?
Bar-Dov is co-founder and chief executive of Heka Global, a company that runs across New York and Tel Aviv and turns the open internet into something banks can actually act on. Heka extracts and structures every verifiable data point it can find on the public web, then hands financial institutions a real-time digital identity for a given consumer - audit-ready, and explainable enough to survive a compliance review.
The customers are not hobbyists. Heka's engine sits inside banks, payment processors, insurers, and pension funds, filling the blind spots in fraud mitigation, credit decisions, and account recovery. Names attached to the work include Barclays, the Pictet Group, Clal Insurance, and ZEDRA. In one deployment with a global payment processor, Heka caught 65% of account-takeover losses without tripping up legitimate customers - the metric that matters most, because the easy way to stop fraud is to also stop everyone else.
In July 2025 the company closed a $14 million Series A led by Windare Ventures, with Barclays and Switzerland's Corner Banca joining in. The money is pointed at the United States, with deeper roots in the UK and Europe to follow.
The credit bureaus were built for another era. Today, both consumers and risk live online.Idan Bar-Dov · Co-Founder & CEO, Heka Global
Heka did not begin as a weapon against fraud. It began, during the COVID lockdown, as a way to reunite ordinary people with their own lost assets - dormant pensions, old savings accounts, money sitting in the gaps where institutional records had gone stale. The trick was tracing a person through their digital breadcrumbs when the official paperwork had stopped keeping up.
That same technique - following someone's verifiable trail across the web - turned out to be exactly what banks needed pointed in the other direction. If you can find a person who has been lost, you can also tell whether a person is real. The asset-recovery startup quietly became a fraud-detection and identity platform without changing its core skill.
Bar-Dov came to it from law, not engineering. He spent his early career as a fintech and high-tech lawyer at international firms, advising the kinds of companies he would later join the ranks of. Founding a startup mid-pandemic was the pivot - from the person who reads the contracts to the person who signs them.
The founding bench is its own story. Alongside Bar-Dov sit Ishay Horowitz, a senior officer out of the Israeli intelligence community, and Rafael Berber, a former global head of equity trading at Merrill Lynch. A lawyer, a spy, and a Wall Street trader - the three lenses a fraud problem actually needs.
The engine surfaces alias use, reputational exposure, and behavioral anomalies across the open web - the tells of synthetic identities, burner accounts, and account takeovers that static internal data never sees.
For credit and insurance underwriting, Heka enriches thin files with external, explainable signals - so a decision can be made faster and still be defended later.
The original mission, still running: tracing missing pension members and reconnecting people with dormant accounts, including overseas members where traditional data sources fall short.
AI has given everyone - both good and bad actors - a significant boost in their day-to-day capabilities.Idan Bar-Dov · On the rise of fraud-as-a-service
Bar-Dov's pitch rests on a shift he describes bluntly: what used to require manual effort by individuals has become organized, global operation, fueled by AI and machine learning. The high-end attack vectors that once belonged to sophisticated syndicates are now available to anyone with an internet connection.
The numbers he points to are uncomfortable. Consumer fraud losses hit $12.5 billion last year, a 38% jump. And by one figure he cites, 34% of consumers have been offered a chance to participate in fraud themselves. Large language and image models now let bad actors replicate personas and mint synthetic identities at scale.
His counter is not more friction for honest users. It is better evidence - real-time, drawn from the same public web the fraudsters operate in, and explainable enough that a regulator, an underwriter, or an auditor can follow the reasoning. Always-on, accurate, explainable: the three words he keeps returning to.