01
Compressed loops
A prototype that used to take a quarter takes a week. We stopped planning around the old number, and we stopped pretending a document can answer a question that a working version answers in days.
About
We work across brand, digital, growth, learning, and research. Each discipline is considered as part of a larger whole, with the same team shaping the work and remaining accountable for how it performs in the real world.
Founder
For most of the last twenty years the cost of finding out whether something worked was high enough that organisations planned instead. Roadmaps, discovery phases, steering committees. That was a rational response to a first version taking two quarters. It is not a rational response to one taking two weeks.
M3V is built for the second situation. Small teams, short loops, real software in front of real people quickly — and then the unglamorous engineering that decides whether any of it is still standing in year five.
The second half is the part people underestimate. Speed is easy to buy now. Judgement about what deserves to survive is not.
Manish Poduval, Founder
Speed
AI did not make software cheaper so much as it made being wrong cheaper. That changes what a small company can reasonably attempt & it changes almost nothing about what it takes to keep the result running.
01
A prototype that used to take a quarter takes a week. We stopped planning around the old number, and we stopped pretending a document can answer a question that a working version answers in days.
02
One small team carries a product from the first conversation to production. Handovers are where speed and understanding are both lost, so we have as few of them as the work allows.
03
Models do the genuinely mechanical parts — scaffolding, migrations, content operations, retrieval, test surfaces. Human judgement stays on the parts that decide whether the thing is any good. We are not an AI agency; we are a company that would be slower without it.
04
The point of speed is not shipping more. It is being wrong in week two rather than in quarter three, while changing your mind still costs almost nothing.
05
Once something is worth keeping it gets the boring engineering: performance budgets, tests, monitoring, backups, a deployment process a team can follow. Blitz speed to find the thing. Engineering discipline to keep it.
Some of that capability goes into client work. Some of it goes into products and research of our own, which is usually where the next capability comes from.
What we do
01
Websites and web software for organisations where the site is business-critical: design, engineering, security, accessibility, compliance implementation and long-term maintenance.
02
Strategy, identity systems and art direction — including the digital brand systems that decide whether an identity survives its first year of real use.
03
Paid media, campaign strategy, creative systems and the analytics needed to know whether any of it worked.
04
Applied investigation into language models, retrieval, drones, security and logistics, mostly with no client attached.
05
Technology education and corporate upskilling, delivered through OpenBootcamp.
How we work
We would rather do a small number of things at a standard we can defend than fill a portfolio. Most of the value in a project is decided by what gets left out.
A website is not finished at launch; it is only beginning to be used. We design for the fifth year as much as for the launch announcement.
We implement against the security, accessibility and governance requirements our clients are held to, and we document what was done. We do not claim certifications we do not hold.
Learning
Knowledge generated through client work and research does not have to stay inside the company. Much of it is more useful as teaching material than as an internal document.
OpenBootcamp is M3V’s technology education platform offering practical, project-driven programmes in AI, software, product and design.
Explore OpenBootcamp