AI systems & software engineering

Your business,
free to grow.

  • Agents
  • Automations
  • Integrations
  • Computer vision
  • Custom ML models
  • Security
  • Observability
  • AI training for your team
  • Voice agents
  • CRM & workflows
  • Document processing
  • Product engineering
  • Technical direction
Let's talk
AI systemsConnected dataCustom platformsReliable deliveryPeople in control
Your businessResearch · engineer · evaluateSystems in use
01

Intelligence that can work inside your business.

An assistant needs context, useful tools and limits. We research the task, connect the right knowledge, and evaluate its decisions before it gets permission to act.

Begin

01 / Our work

Case study: from idea to production.

We start with what the business wants to achieve. For TerapieAcasă, that meant people could find a therapist, book a session and attend it from home.

See more of our work

TerapieAcasă / Project film

TerapieAcasă · A platform we built

A walkthrough of the platform’s features, with illustrated interfaces and sample data.

Read the story

A match before an account. Preferences pass through filters and weighted scoring to a therapist introduction.

Keep the conversation close. Real-time messages keep the client and therapist connected between sessions.

Meet from home. Video sessions belong inside the same relationship and workspace.

The appointment stays in sync. Booking and Google Calendar sync keep the working day connected.

Help that knows the platform. AI assistance retrieves platform guidance and keeps sources close to the answer.

Give the operator a clear view. Client records, appointments and follow-ups meet in the CRM.

The relationship carries through. Matching, conversation, booking and video connect to the records behind the service.

A booking sets the next steps in motion. Calendar updates, confirmation email and client records follow the appointment.

A session becomes a record. Payments, accounting and automated follow-ups support the business.

In conversation / TerapieAcasă

The platform, from the CEO’s seat.

Case study 1

Flagship / In production

TerapieAcasă

An online therapy business, engineered end to end.

We turned an idea for online therapy into a service that people can use from their first visit to their session. The operator manages the exceptions; the platform carries the routine work of matching, appointments and payments.

View project
Let's talk

02 / Make room

The shape is already there.
We help it emerge.

Your people know the work. We sit beside them, find what keeps getting in the way, and build from there.

  1. 01

    Understand

    Follow the work across the business: its people, decisions, knowledge and systems, and what they need to make possible.

  2. 02

    Remove

    Design and test the system around that work. Keep judgement and approval with the people who own them.

  3. 03

    Release

    Put it into use together. Train the team, learn from the difficult cases, and keep improving the work.

See how we work
After Michelangelo · removing what stands in the way

The people using it

Build capability into the team.

A system improves through the people working with it. We prepare training around your workflows, work through real cases together, and leave materials your team can return to as the tools change.

How we work alongside your team

03 / More of our work

Selected platforms, systems and people.

Explore the work

Built across the stack · Shared with the community

BodyOSCOMMUNITY EDITION

From a body in motion to a product you can make your own.

BodyOS brings together a web workspace, a mobile training app, native computer vision, a data foundation and machine learning research. We built it. Now we’re sharing the source.

Explore the live demoGet BodyOSThe project and release

Case study 2

Product engineering

Creators Multiverse

Generate, edit and publish across platforms.

We built media generation, integrated video and image editors, SEO analysis and cross-platform content generation in one connected workspace.

View project

Case study 3

Teaching / 100+ people

The Art of Programming

Make the tools understandable enough to use independently.

Python and AI taught through things people build themselves. That same habit shapes a handover: your team should understand the system they are taking on.

View project

04 / Under the surface

You should be able to see how it works.

A useful system comes with controls, a record of its decisions, and someone who knows what to do when it fails.

01

Safety

Decide what can run automatically, what needs approval and how a person stops it.

02

Observability

Follow a run from its input to its result, including the decisions and failures along the way.

03

Compliance

Agree which data is needed, who can access it and how long it stays.

Explore a sample project: follow the work, review a decision and meet the people responsible.

Explore the workbench

Tools we build and run with

  • Python
  • TypeScript
  • Go
  • Rust
  • SQL
  • React
  • Next.js
  • FastAPI
  • Node.js
  • Django
  • PostgreSQL
  • Supabase
  • MongoDB
  • BigQuery
  • dbt
  • Kafka
  • Vertex AI
  • TensorFlow
  • scikit-learn
  • LangGraph
  • LangChain
  • OpenAI
  • Anthropic Claude
  • Google Gemini
  • Stripe
  • Remotion
  • Playwright
  • Deno
  • Expo
  • Google Cloud
  • GKE
  • Cloud Run
  • App Engine
  • Cloud SQL
  • Kubernetes
  • Helm
  • Docker
  • Terraform
  • Flux
  • Argo CD
  • Argo Workflows
  • Atlantis
  • GitLab CI
  • GitHub Actions
  • Ansible
  • Datadog
  • Grafana
  • Istio
  • gRPC
  • Keycloak
  • Resend
  • Meta Ads API
  • Cisco
  • MikroTik
  • Linux

Chosen for each project, from a longer bench.

The next piece of work

What would you like to make possible?

Tell us about the system you need, the operation you want to improve, or the idea ready to become a product.

Let's talk