01
Production is the only honest measure.
If it does not run in the real world, it does not count.
Clinical grid. Fresh logic. Work that survives production.
Where design, engineering, and AI stop being three separate conversations.
We are a collective of founders, engineers, and designers. We build software that works the way people think.
01 / Build Thinking
Your product works.
It just doesn't work the way it should yet.
02 / Delivery
Not handed between separate teams. Not dressed up in process theatre. Built directly.
01
Who We Are
We started as a group of people who kept finding each other on the same hard problems, founders who had built and shipped, engineers who had seen good ideas fail because the execution was ordinary, designers who were tired of being handed a brief after all the real decisions were made.
At some point the informal became intentional. This is what that looks like.
We work at the intersection of product design, engineering, and AI. Not as separate practices handed between teams, as one continuous act of building something worth building. We take twelve clients at a time, work directly, and leave with nothing held back.
Operating Rule
Small enough to stay inside the real problem.
Engagement Shape
12
Clients max
1
Continuous team
0
Held back
02
What We Believe
01
If it does not run in the real world, it does not count.
02
The best work we have done involved arguing successfully against our own ideas. Honesty is more useful than agreement.
03
Past twelve clients, the people who know the problem stop being the people who solve it. We have seen that version. We do not want to build it.
Useful honesty
We have told clients their AI idea was a distraction. The ones who listened are glad we said it.
Structural choice
Small is not a posture. It is how the people who understand the problem remain the people solving it.
What we build
We design products, set up the infrastructure that lets them scale, and implement AI where it belongs: inside the workflow, inside the interface, and inside the operational reality of the business.
Lane 01 / Product, systems, infrastructure
01 / Product Design
For SaaS products that need sharper flows, clearer hierarchy, and a product surface that finally matches the ambition underneath it.
02 / Scaling Systems
Component libraries, token architecture, coded React components, and the foundations that let a product scale without getting inconsistent.
03 / Platform Engineering
Architecture, platform setup, internal tooling, and the engineering decisions that keep product velocity from collapsing under growth.
04 / Custom AI
Custom AI and machine learning systems for classification, prediction, retrieval, and workflows that off-the-shelf models get wrong.
01 / Product Design
For SaaS products that need sharper flows, clearer hierarchy, and a product surface that finally matches the ambition underneath it.
02 / Scaling Systems
Component libraries, token architecture, coded React components, and the foundations that let a product scale without getting inconsistent.
03 / Platform Engineering
Architecture, platform setup, internal tooling, and the engineering decisions that keep product velocity from collapsing under growth.
04 / Custom AI
Custom AI and machine learning systems for classification, prediction, retrieval, and workflows that off-the-shelf models get wrong.
Lane 02 / Automation, interfaces, deployment
05 / Workflow Automation
Operational automations and agentic flows that remove repetitive manual work across teams, tools, and approval chains.
06 / Embedded Intelligence
Classification, summarisation, routing, and next-action systems integrated where users already work, not hidden behind a separate AI tab.
07 / Natural Language
Interfaces on top of your existing data so teams can ask questions, trigger actions, and get answers without SQL or analyst bottlenecks.
08 / Deployment Surfaces
Operational AI delivered inside the interfaces people actually use, including WhatsApp flows and systems built for Indian language usage.
05 / Workflow Automation
Operational automations and agentic flows that remove repetitive manual work across teams, tools, and approval chains.
06 / Embedded Intelligence
Classification, summarisation, routing, and next-action systems integrated where users already work, not hidden behind a separate AI tab.
07 / Natural Language
Interfaces on top of your existing data so teams can ask questions, trigger actions, and get answers without SQL or analyst bottlenecks.
08 / Deployment Surfaces
Operational AI delivered inside the interfaces people actually use, including WhatsApp flows and systems built for Indian language usage.
03
What We Build
Product design, enterprise software, and AI built together, not sequenced through separate departments.

03A
In-House Products
Built Inside
Things we built because they did not exist the way we needed them to.
Not client work. Not prototypes. Products we use, refine, and put into the world for anyone to try.
Why this exists
All of these are live. Some are open source. All of them started as something we needed and could not find. If you are building in a similar space, start here.
04
The Way In
We look at the product, the data, and the real constraints. We produce a specific, technical, prioritised roadmap. At the end of two weeks, the decision is yours. Work with us, work with someone else, work with your own team. The document belongs to you regardless.
Observe
Product, workflows, and failure points mapped fast.
Diagnose
Technical, design, and AI constraints prioritised in sequence.
Decide
You leave with the roadmap whether or not we continue.
Outcome
A technical roadmap with enough specificity to act on, not a vague strategy deck.