Skip to content

Services

A technology partner for selected projects.

We do not take on every brief. We focus on work that sits close to the infrastructure behind our own products, so we bring experience rather than guesses.
01

SaaS and custom software development

Web applications, subscription platforms, dashboards and internal tools.

What we build

  • Designing and building a first release
  • Growing an application without rewriting it
  • User management, permissions and billing flows
  • Admin dashboards and reporting screens
Who it is for
Startups growing a product, and companies moving their operations into software.
The problem we solve
Processes held together with scattered tools stop scaling. Teams start doing the same work in different places.
What we deliver
A working, maintainable application delivered with its source code, setup steps and data model documentation.
Technology
Next.jsTypeScriptReactNode.jsNestJSGoPostgreSQL
02

API engineering and integration

REST APIs, backend services, authentication and integrations with outside systems.

What we build

  • API design and versioning from scratch
  • Authentication, authorisation and rate limiting
  • Integration layers for outside services
  • API documentation and developer experience
  • Performance measurement and query optimisation
Who it is for
Teams whose product has to talk to other systems, and companies opening their own API to the outside.
The problem we solve
Unreliable integrations slow down the whole product. An undocumented interface that is hard to debug locks the team in place.
What we deliver
A documented, tested and observable API layer, delivered with its endpoint list, error shapes and example requests.
Evidence from our own products
The API Deposu catalogue runs 443 API sources, our own validation services and an MCP server on one platform.
Technology
Node.jsNestJSGoPostgreSQLRedisOpenAPI
03

AI integration and automation

Adding AI features to existing products, document processing and workflow automation.

What we build

  • Assistants and search experiences embedded in a product
  • Question answering systems that run on your own data
  • Agents that call tools to reach outside systems
  • Automation that processes documents, forms and email
  • Taking repetitive manual work out of the loop
Who it is for
Companies that want manual work automated, and teams adding an AI feature to their product.
The problem we solve
AI experiments often stay disconnected from the product. The real work is connecting the model to the right data and the right flow.
What we deliver
A measurable automation, or an AI feature inside the product, with its cost, failure modes and limits defined.
Technology
LLM APIsRAGTool callingVector searchpgvectorTypeScript
04

Data engineering and geospatial solutions

Data collection and transformation, and applications that run on a map.

What we build

  • Collecting and cleaning data from scattered sources
  • Scheduled data refresh pipelines
  • Geospatial querying and analysis on PostGIS
  • Transport and movement data processing
  • Map interfaces and location based services
Who it is for
Teams working with location, logistics, transport and geographic data.
The problem we solve
Scattered, dirty data makes decisions harder and software slower to build. Showing it on a map is a separate engineering problem.
What we deliver
A queryable, verified data layer and the application that uses it, with sources and refresh frequency documented.
Evidence from our own products
İstanbulaşım processes real GTFS data and aircraft tracking data on a live map.
Technology
PostgreSQLPostGISPythonGTFSNode.jsMapLibre

Complementary capabilities

These areas support our main services. Depending on the project, one of them can become the deciding factor.

Mobile applications

Applications built from a shared codebase for iOS and Android.

Cloud and infrastructure

Deployment, environment management, monitoring and release pipelines.

MVP development

Turning an idea into a working release that real users can test.

Product architecture

Writing down technical decisions and the data model, then prototyping them.

How we work

Four steps from idea to production.

Every project follows the same order. You always know what we are working on and what comes next.

  1. 01

    Discover

    We pin down the problem, the user and the constraints, then write the scope down. That includes what we will not do.

  2. 02

    Design and architecture

    We work out the interface flows, the data model and the technical decisions, and test the risky assumptions early with small prototypes.

  3. 03

    Build and integrate

    We deliver working software in slices. Each slice arrives tested, reviewed and ready to run.

  4. 04

    Launch and improve

    We launch it, measure it and adjust it against real usage. We stay reachable for maintenance after handover.

Ways of working

We do not share a price before the scope is clear. Once it is written down, we propose either a fixed price or a monthly engagement.

Project based

Written scope, staged delivery, handed over at the end.

Work with a clear start and end.

Monthly engagement

An ongoing model where we set the priorities together each month.

Teams with a live product that need steady development capacity.

Takeover

We review an existing codebase, document the risks, then take over maintenance.

Projects whose developer has left, or whose pace has stalled.

Common questions

Which technologies do you use?

Mostly TypeScript and Node.js, with React and Next.js on the interface side and PostgreSQL with PostGIS in the data layer. Go comes in where a service needs to be fast. We pick the technology for the project, not out of habit.

How do you price projects?

We do not quote before the scope is written down. Once it is clear we propose either a fixed project price or a monthly engagement. The first call and the scoping work are free.

How long does a project take?

Duration depends entirely on scope, so we do not quote a number up front. After scoping we share a staged plan and a date range for each stage. We prefer to deliver a first working release early and build on top of it.

Can you work alongside our own team?

Yes. We can work next to your team, contribute to your codebase or take ownership of a specific module. We agree on code review, release process and communication channels before we start.

When does adding an AI feature make sense?

When there is a user flow that needs a decision and the outcome can be measured. We do not add AI because it is fashionable. Some problems are best solved with rules, others with a model, and the scoping work is where we settle which is which.

Will you take over an unfinished project?

Yes, after a review. We read the codebase and share the technical debt, the risks and the takeover scope in writing. In some cases that review has led us to advise against continuing at all.

My idea is still vague. Is it worth talking?

Yes. We are used to early ideas, our own products started that way. The first call is about finding the outcome and the constraints, and working out which problem actually needs solving.

Do you reuse your product infrastructure in client work?

We reuse the approach and some components, but we do not resell a product as a service. Each project has its own data model and flows, so the solution is built for the project.

Projects start with a conversation.

In the first call we work out the scope, the technical risks and the right approach. We do not quote yet. We understand the problem first.

Discuss a project