Seven products. Every one started as our own problem.

Calls the lead in seconds, then proves which ad paid for it

A console that phones a Meta or Google lead within seconds of the form submission and qualifies them against rules the operator wrote in plain language. If they qualify, it books the appointment. When the sale closes, it traces the revenue back to the exact ad and creative that produced it, then reports that conversion to the ad platforms so they optimise on revenue instead of form fills.

• The call goes out while the form confirmation is still on the lead's screen
• Plain-language qualification rules the operator edits themselves
• Closed-loop attribution back to Meta and Google conversions APIs
• Four surfaces: watch the money, work the leads, run campaigns, prove the maths

For

Anyone buying leads on Meta or Google who cannot say which ad produced the last closed sale.

Why we built it

Every tool in this space solves one half: a dialler that knows nothing about attribution, or an attribution product that cannot pick up a phone. Closing the loop is the whole point, so it had to be one system.

Local-first screen recorder and automatic-zoom editor

Record your browser and the clicks turn into smooth camera moves. Every one of them stays editable, and the footage never leaves the machine. A Next.js product plus a Chromium extension that captures screen, microphone, camera, cursor and interactions.

• Click-aware automatic zooms and follow-cursor pans
• Screen, camera, microphone, captions and effects on one timeline, every zoom still editable
• Local MP4/WebM export — no media upload step
• One-time $119 lifetime price across four devices, not a subscription

For

Founders, developer-advocates and solo marketers who ship their own demo videos.

Why we built it

We were making demo videos for our own launches and paying a subscription to upload our footage to somebody else's server to do it. Local-first was the point.

A trust score for handwritten property documents in the DRC

A French-first property marketplace for the Democratic Republic of the Congo. Titra reads a property document — typed or, as is common in the DRC, handwritten — pulls out what is actually on the page, and applies a versioned rule set to produce an explainable trust score. A score, not a certificate.

• Bilingual FR/EN, French first, resolved server-side so the first byte is already in the right language
• Versioned scoring rules — the badge comes from the score, not from the listing copy
• Decision support for a human, explicitly not a legal certification
• No authoritative DRC land registry is connected, so 10 of the 100 points are permanently outstanding, and the product says so

For

Buyers, sellers and the notaries, agents and inspectors between them, in Kinshasa first.

Why we built it

The interesting problem is not the marketplace, it is producing a score a human can argue with. Every point is traceable to a named rule and a line in the document.

Matchmaking for the room you are already in

Guests answer a short question set, the product matches everyone in the room, and the reveal lands at the same second for everybody. Built for house parties, singles nights and team offsites — ten-minute setup, no app to install, phone numbers masked on every plan.

• Event templates that pre-fill mode, question set and reveal timing
• AI question suggester with a workplace-vetted bank for professional events
• Door check-in so no-shows never pollute the match pool
• Free up to 50 guests, $99 per event beyond that

For

Party hosts, community organisers and teams running an offsite.

Why we built it

We had already built a 50,000-user events platform for someone else. This is the small, sharp version of the part that actually made those events work.

The back office behind Offline by Happy Hour

The operations system that runs the Offline by Happy Hour business. Bookings, traveller records, CRM and pipeline, itineraries, invoicing, receipts, payments, tax handling and finance reporting sit in one console, so the answer to any question about a trip is in one place rather than five files.

• Bookings, travellers, trips and itineraries as one data model
• Invoice and receipt generation as real PDFs, not screenshots
• Payments, international tax handling and finance reporting
• Built for the same team behind the Shark Tank India platform we shipped

For

The operations team behind Offline by Happy Hour.

Why we built it

Nothing off the shelf modelled bookings, travellers, itineraries, invoices and international tax as one thing. Buying four tools and reconciling them by hand is what we were replacing.

Over-the-air deployment for React Native

Ship a React Native update without waiting on an app store review. A deployment platform the studio built and maintains for its own mobile work first.

• A JavaScript-level fix reaches installed apps without a new store submission
• Our own React Native releases run on it, which is why it is still maintained

For

Teams shipping React Native apps, starting with ours.

Why we built it

We hit the problem on our own mobile releases first. Building it was cheaper than living with the review queue.

AI-assisted audience growth, running on our own accounts first

An AI-assisted social growth product the studio builds and runs on its own accounts. It is live and it is early. It is listed here because a studio that claims to operate its own products should list all of them, including the small ones.

• We post through it ourselves, which is the only reason it keeps getting fixed

For

Founders and operators building an audience alongside a product.

Why we built it

A studio that tells clients to measure what they ship should be running its own growth on something it can see inside.