Product development company · Atlanta & Chennai
Proof that a person is real. Structure that keeps information findable. Four platforms, built and owned in-house.
Portfolio
Tilde ID is the core: an identity platform that everything else can eventually stand on. Yoomark has been running since 2016 and has published over two million ads. Tubelx and NoticeB are new. What they share is a reading of where value moves when generation becomes free — towards proof that a person is real, and structure that keeps information findable.
Core platform · Tilde ID
Tilde ID is the platform the rest of the company is organised around. A person enrols their face once — envelope-encrypted as a template, with raw camera frames never leaving the device — and that single enrolment then works at the front desk, at the login, in a meeting, on a signature, and onward across an ecosystem of 11 surfaces. Three are live today.
Why it compounds: every new surface makes the original enrolment more valuable, and none of them requires the person to enrol again. That is the difference between a feature and a platform — and it is why identity is the one we lead with.
Documents signed by a biometrically verified human
In buildNext surfaces
Patient identity and check-in
Banking and KYC
Campus access and attendance
Authority for AI agents acting for a person
Fraud-resistant checkout
Civic identity
Private age proof
Why these problems
When anything can be generated, the scarce things are proof and structure.
Producing a video, a profile or a thousand plausible accounts is now close to free. Everything downstream of that is an open problem: how a service knows a user is human, where generated media gets made and watched, how a community keeps a usable record, how a real local business still gets found. Our four platforms each take one of those.
Verification is becoming mandatory across finance, marketplaces and public services at the same moment biometric databases became the thing regulators fine you for. The answer is matching without retention — and Tilde ID is in beta at both the door and the login.
Models are commoditising: a creator can already reach eight of them through one studio. What has not been solved is the gap between making a video and having anyone watch it. Tubelx closes that gap by owning both ends.
A feed shows what is newest and discards the rest. When content is infinite, that is a worse trade every year. Following a board rather than a person puts the reader back in charge of what persists.
Local search is dominated by platforms that sell a business its own customers back to it. A directory where placement is the product, and the targeting is what someone is actively looking for, is a cleaner trade on both sides — and one we have now run for a decade.
How we work
We hold equity in the platforms we build and we run them ourselves. That changes every decision that follows.
The best time to build a replacement is when the incumbent approach is becoming illegal, uneconomic or obsolete — whether the force doing that is regulation or a model release. We look for markets where the direction of travel is already decided and the product does not exist yet.
Most product failures are not distribution failures — they are a technical assumption that turned out to be false. We prove the biometric matching, the ranking model, the sync layer before we build a landing page for it.
Every platform runs live, with real users, well before it is finished. A product that has never met production has not been tested; it has been imagined.
Identity, storage, deployment and observability are built once and used by everything in the portfolio. The structure only earns itself if the second product costs less to build than the first.
Notes
Occasional writing on identity, discovery and the engineering behind them.
Why the safest place to put a fingerprint is nowhere, and what that constraint does to the architecture of an identity system.
On 21 May 2025 we started building. A note on what we set out to do, and the problems we chose to start with.
Start a conversation
The conversations we find useful are quite specific: investment, distribution and integration partnerships, and people who want to build these products with us. If you are on that list, we would like to hear the specifics.