An engineering team from Lviv: medical devices, rail transport, telecom and IoT. From architecture and circuit design through to firmware, cloud and automated testing.
The areas our engineers have worked in commercially — from device firmware to the dashboard that controls it.
Firmware for microcontroller and embedded Linux devices: power management, tight memory budgets, peripheral drivers, over-the-air updates. We write for bare metal, RTOS and Yocto builds.
Development under safety requirements: software classification up to IEC 62304 class C, requirement traceability, integration with hospital services and consumables tracking.
Distributed fleets of hundreds of devices: central control, schedules and playlists, telemetry collection, status dashboards and remote updates. Our own product ADSYSTEM runs on this stack.
Software for switches, routers and access points: configuration tools, protocol implementations, CLI and SNMP engines, northbound APIs for SDN systems, DPDK traffic generators.
Real-time systems for trains and cars: CAN communication, integration with on-board subsystems, motor control, map data for autonomous driving.
We build test frameworks and hardware simulators so firmware can be verified without physical devices: accelerated clocks, parallel runs, sensor self-tests, CI on every commit.
We design the system before any code is written: HLD and DLD, scope and budget estimates, component selection, protocol security. We can take the hardware too — from schematic to board.
We pick proven tools over fashionable ones: the device will run for years, and someone will have to maintain it — possibly another team.
You see a working version long before launch. We build the hardware simulator up front so nobody waits on the board.
One hour: what the device or system is, hardware constraints, safety requirements and deadlines.
1–2 daysHLD, platform and protocol choices, a written estimate of scope, budget and risks.
up to 5 daysWe build a hardware simulator and prototype the key components — changes are cheap at this stage.
2–4 weeksA demo every two weeks on a test bench, automated tests and CI on every commit.
from 6 weeksHandover of code and documentation, remote updates, monitoring and a roadmap.
ongoingProjects our engineers worked on in various roles, from architect to developer. Client names stay private under NDA, so we give the industry, the task and the result.
A distributed Raspberry Pi system for indoor advertising: content is set in one dashboard and rolls out to every screen, with device health and playback statistics visible in one place. The hard constraint was minimum cost per device on site.
Real-time software architecture for a new train: integration with on-board subsystems and hardware, low-level CAN optimisation, a web interface for configuration and status monitoring, and hardware simulation for testing.
The device tracks every sterilisation cycle: duration, instrument, the physician who started it and the detergents used. Data flows into hospital services, and the device reorders consumables on its own. Secured communication and remote firmware updates.
A test platform for embedded medical device software: a C++ core scripted through Google V8, plus simulated wireless modules and BLE chips that talk to each other. Scenarios run without any physical device on the bench.
Configuration tooling for VLAN, STP, LLDP, LAG and LACP, an SNMP and CLI engine, and a northbound API for an SDN system. A separate track covered VoIP: SIP, H323, SS7 and IMS signalling for carrier switches.
The six ways an outsourced software project usually goes wrong — and what we do so it does not happen on yours.
“We will only see the result at the end, when nothing can be changed.”
A working build on a test bench every two weeks, starting with the first sprint. You set the priorities for the next one, so a change of direction between demos costs nothing.
“The estimate will double somewhere in the middle.”
Scope, budget and risks arrive in writing before development starts. Anything new goes through a re-estimate you approve first — it never shows up as a silent overrun on the invoice.
“The board is not ready, so the software will sit and wait.”
We build hardware simulators up front, so firmware and the dashboard are written and tested in parallel with the board. A test framework built this way is used by 80+ engineers on a medical device programme.
“We will be locked into this vendor forever.”
The code sits in your repository from day one; HLD, DLD and build instructions are part of delivery, not an extra. Any other team can pick the project up, and that is written into the contract.
“One key developer leaves and the project dies.”
No project runs on a single person: an architect owns the decisions, engineers review each other’s code, and documentation is written as the work goes, not afterwards. Teams of up to 60 people have been run this way.
“Our field has requirements an outside team will not understand.”
Medical devices up to IEC 62304 class C, real-time rail systems, carrier telecom, network equipment: 35+ delivered projects across 11 countries. We start from your constraints and standards, not from a template.
Most clients start small: an audit or a discovery phase first, a team only once the scope is known. Each step is useful on its own, and you are free to stop after any of them.
An independent assessment of a system you already have, or of an architecture you are about to build.
A fixed-price phase that turns an idea into a specification, a plan and a firm number.
Assigned engineers work in sprints against your priorities.
Most of this work was delivered under non-disclosure agreements, so we describe the industry, the task and the result, but not the company. We can go into detail in a meeting under an NDA on your side.
The price depends on constraints: whether safety requirements such as IEC 62304 apply, how many hardware variants need support, and whether simulators and certification are involved. An architecture review is €2,500, a fixed-price discovery phase €4,750, and a dedicated team starts at €65 per engineer hour. The first call and a rough scope estimate cost nothing.
No. Around the device we also build what the user sees: control dashboards, telemetry views, backend and cloud. Our own product ADSYSTEM is exactly that — player firmware plus a dashboard with playlists and statistics.
Yes. We start with an audit: state of the code and documentation, risks, cost of support. From there we either take it on or lay out a staged modernisation plan. The team has picked up legacy codebases before.
You do. The code sits in your repository from day one and rights transfer under the contract. HLD, DLD and the rest of the project documentation are part of delivery, so any other team can pick the work up.
Remotely. Our engineers have delivered projects for clients in the USA, the UK, Germany, the Netherlands, Norway and South Korea. Working language is English at Advanced level, meetings online, reporting in whatever format suits you.
We reply within one business day. The first consultation and scope estimate are free, with no strings attached.