ventures,
[r]ethought.
Seasoned operators obsessed with problems worth solving. AI gives us speed. Experience gives us judgment.
Tech execution is no longer the hard part. AI solved that.
The bottleneck is now knowing what to build and having the grit to sell it. We talk to customers before we write a line of code. We kill what isn't working before anyone gets attached. We do things that don't scale until we find the signal worth scaling.
Every company we build is AI-native from Day 0. Not AI as a feature. AI as the foundation. And every venture makes the next one sharper.
from signal to scale.
Four phases. Each earns the right to the next.
Most ideas don't survive phase three, and that's how the studio gets smarter.
Find the fight worth fighting.
AI absorbs market signals at a pace no analyst team can match: competitors, regulation shifts, pricing patterns, user sentiment. It surfaces connections humans would miss. But seeing a pattern isn't the same as choosing a battle. Operators decide where timing, pain, and capability intersect. That's taste, not data.
Collapse idea to evidence.
The gap between "I wonder if..." and "let's find out" used to be weeks. Now it's hours. AI generates code, scaffolds experiments, and builds testable prototypes while operators own the architecture and the question behind it. Don't polish. Get it in front of real people and see what happens.
Listen harder than you build.
This is the most human phase. Manual onboarding. Hand-written emails. Sitting across from a customer and hearing what they're not saying. AI synthesises the signals (clusters objections, spots patterns in feedback) but humans feel the signal first. No dashboards. Just conversations that change the product.
Double down or walk away.
The studio's edge is optionality. Validated pull gets capital, team, and infrastructure. Everything else gets killed fast, before sunk-cost thinking sets in. Every shutdown feeds learnings back into the engine: reusable code, customer research, operational playbooks. Nothing is wasted. The next venture starts from a higher baseline.
obsessed with a problem worth solving?
Talk to ussmall crew. compounding output.
Headcount is not a proxy for ambition. We run AI agents across every stage of the venture lifecycle: research, prototyping, validation, distribution. We keep a tight crew of operators focused on the decisions machines can't make.
AI handles throughput. Humans handle judgment. The result: ventures that get our full commitment, each compounding what we know. Constraints sharpen us. Speed compounds.
what we're building.
We go deep, not wide.
Every venture gets our full conviction, until the market tells us otherwise.
Invest in our pipeline
deetech Ranks #1 In Audio Deepfake Detection
An independent benchmark put deetech first among 23 systems, with 99.56% accuracy. The bigger win is generalisation: detection built for the next deepfake, not the last one.
Read article
The Phone Has No Liveness Check: Why We Created BotBlock
Cloned voices and scam bots now call contact centres, and the phone has no liveness check. BotBlock stops voice scams before hello.
Read articlewho's behind this.
Mind, heart, and soul in every venture we build.
PhD candidate in AI. Spent a decade in strategy consulting, analytics, and M&A before building [r]think. Now obsessed with turning operator judgment into scalable products.
We're looking for someone who knows their industry cold and has the commercial grit to sell before the product is polished. Deep sector expertise, a nose for real pain points, and the endurance to grind through early revenue. You've closed deals, not just decks.