Software studio · Tallinn, Estonia

We build the software companies run on.

Operations platforms, client portals and internal tools. We scope the workflow, build the system, and hand it over as code you own. Everything we ship runs in production.

Start a projectSee the work→

Selected work

Systems we designed, built, and still run.

Not prototypes or pilots — production software that people log into every morning. A selection of what we've built and continue to operate.

01

PRDC Platform

Multi-tenant SaaS · property & facility operations

PRDC/Tickets

Tickets126 open
New ticket
RefIssueStatus
TCK-4192Lift stuck between floors 3–4Dispatched
TCK-4191Heating loop pressure dropIn progress
TCK-4188Garage door sensor faultAwaiting parts
TCK-4186Quarterly ventilation serviceScheduled
TCK-4180Lobby lighting replacementResolved

Interface shown with illustrative data. No customer record appears on this page.

One place for a property manager to take a problem in, turn it into work, assign it, and bill for it.

A tenant reports by Telegram or web. The system triages the report, opens a ticket, starts the SLA clock, dispatches a technician, tracks their hours, and closes the loop into the contract and the invoice. Thirty-plus modules share one permission model and one activity log, so each workspace switches on only the parts it needs.

Built with

  • Next.js
  • TypeScript
  • Prisma
  • PostgreSQL
  • Redis
  • Telegram Bot API

What's in it

Work

  • Tickets
  • Facility tasks
  • Maintenance
  • Projects
  • Reassignments

Commercial

  • Clients
  • Contracts
  • Finance
  • Pipeline
  • Invoicing

People

  • Users & roles
  • Time tracking
  • Leave
  • Resource planning

Intelligence

  • AI assistant
  • RAG search
  • Analytics
  • Activity hub
  • Audit log
02

Commodity trade platform

Client platform · shown anonymised under NDA

Trade/Trips

Open tonnage

4,180t

In transit

17trips

Escrow held

1.24M USD

Awaiting release

3

Trips17 in transit
New trip
TripNet tStatus
TRP-229134.2In transit
TRP-228833.8At border
TRP-228428.6Offloaded
TRP-227935.0Offloaded
TRP-227131.4Weighbridge gap

Interface shown with illustrative data. No customer record appears on this page.

Trucks leave a loading site and money moves against them. This tracks both, and will not let the two drift apart.

Purchase orders break down into trips. Each trip carries weighbridge readings, customs paperwork and photographs, which a vision model reads so nobody retypes a tonnage at midnight. When a load is confirmed the platform releases the matching escrow tranche and raises invoices in both directions — to the buyer for the goods, to the financier for recovered costs.

Built with

  • Next.js
  • TypeScript
  • Prisma
  • PostgreSQL
  • Auth.js
  • Vision LLM

What's in it

Trade

  • Purchase orders
  • Trips & deliveries
  • Products
  • Suppliers

Money

  • Cash calls
  • Escrow disbursement
  • Dual-direction invoicing
  • Payments

Logistics

  • Hauliers
  • Clearing agents
  • Weighbridge reconciliation

Control

  • Document extraction
  • Role-based access
  • Supplier portal
  • Activity log

What we'll build for you.

Most of our work replaces a spreadsheet that quietly became a system, or a process that lives in three people's heads. If you can describe how the work flows, we can scope it.

  • 01

    Operations platforms

    Take work in, route it, track it, close it. Ticketing, dispatch, scheduling, SLA clocks, and an audit trail that proves what happened and when.

  • 02

    Client & supplier portals

    Scoped outside access to the same data your team already works in, so counterparties stop emailing for status and start reading it themselves.

  • 03

    Document & data extraction

    Invoices, delivery notes, weighbridge tickets and scanned PDFs read into structured records — original attached, and a person confirming anything the model was unsure about.

  • 04

    Integrations

    Accounting, CRM, payments, e-signing, mapping, messaging. The systems you already pay for keep working; the new one talks to them.

  • 05

    Chat & bot interfaces

    Telegram, Slack or web, for the people who are never going to log into a dashboard. Field staff send a message; it arrives as structured work.

  • 06

    Reporting & analytics

    The handful of numbers the business actually steers on, on screens that load fast enough to be worth checking daily.

How a project runs.

01

Scope

One to two weeks. We sit with the people who do the work, map the process as it actually runs rather than as the org chart says it does, and write down what is in and what is out.

You get: a written scope, a fixed price, a date

02

Build

Two-week cycles against a staging URL you can open whenever you like. You watch it work while it is being built, which is while changing your mind is still cheap.

You get: working software, every fortnight

03

Handover

Source code, documentation and infrastructure in accounts with your name on them. We stay on for support if you want it, and it costs you nothing if you don't.

You get: the repository, and the keys

No retainers, no seat licence on your own software, and no clause that makes leaving expensive. If we are still working together in year three, it should be because the work is good.

What we won't do.

We build systems people depend on at work. These are the commitments that come with that — written down rather than implied.

Humans stay in control
Software assists; people decide. Anything consequential — a payment released, a contract closed, a technician sent out at three in the morning — keeps a person in the loop, with a clear escalation path for when the automation is unsure.
Your data stays yours
We do not train models on it and we do not sell it. The architecture follows GDPR and the EU AI Act: where a model is involved it sees only what it needs, only for as long as composing the answer takes.
Automation says so
Every bot and assistant we ship identifies itself as software. Nothing we build pretends to be a person, and no automated decision arrives without a record of how it was reached.
Built expecting failure
Monitoring, alerting and correction paths ship with the first release, not after the first outage. We agree what an acceptable error rate looks like before launch, then measure against it.
No dark patterns
No manufactured urgency, no engagement mechanics, no interface that makes cancelling harder than subscribing. Not for any client, at any budget.
Security and accountability
Least-privilege access, encryption in transit and at rest, and a complete audit log. When something goes wrong you get told what happened, when it happened, and what we changed.

Tell us what needs building.

Send a few paragraphs about the workflow you want to replace or the product you want to launch. You will get a written scope with a price and a date, rather than a discovery call that turns into four more.

hello@prdc.ee

Tallinn, Estonia · English and Estonian · Replies within one working day