Mekriva

Custom firmware and real-time software for microcontrollers and microprocessors — from bare-metal C to RTOS-based architectures. We develop full BSP stacks, hardware abstraction layers, peripheral drivers and sensor integration across STM32, ESP32, NRF52840, RP2040, RISC-V and AVR platforms. Whether you need an ultra-low-power sensor node consuming microamps in sleep or a complex multi-core system managing real-time control loops, safety tasks and wireless communications in parallel, we deliver firmware that is reliable, maintainable and production-ready. Our work spans the full embedded software stack: bootloaders, device drivers, middleware, communication protocol implementations and application logic. We use FreeRTOS, Zephyr or bare-metal approaches depending on the constraints — power budget, latency requirements, code size limits or safety standards. We write clean, documented code with testability in mind: unit tests at the driver level, CI pipelines for automated build and test, and documentation sufficient for your team to maintain and evolve the project independently after delivery.

End-to-end electronic design covering the complete development cycle from concept to validated prototype. Our hardware engineering spans analog and digital circuit design: power supply topologies, signal conditioning circuits, ADC front-ends, microcontroller integration, RF layouts and mixed-signal PCB design for production environments. We work in KiCad and Altium, selecting components with availability, longevity and supply chain resilience in mind. PCB layouts are optimised for signal integrity, thermal management and Design for Manufacturability — we deliver Gerbers that go straight to fabrication without surprises. We assemble and bring up prototypes in-house, validating functionality at each stage with oscilloscopes, logic analysers and signal generators. EMC pre-compliance testing, power consumption profiling under real operating conditions and environmental stress screening are standard before any design is considered complete. On the firmware side, we perform the critical bring-up phase: initialising peripherals, tuning power modes, characterising sensor behaviour and writing the low-level drivers that form the foundation of the product's embedded software stack.

Connecting devices, sensors and industrial equipment to broader systems — reliably, efficiently and at the right abstraction level. We specialise in the integration layer between the hardware and the application: protocol implementation, driver development, middleware and gateway design. On the wired side, we implement SPI, I2C, UART, RS-232, RS-485, CAN, CANopen, Modbus RTU/TCP and custom binary protocols, handling edge cases like bus contention, CRC validation, error recovery and hot-plug behaviour correctly from the outset. For wireless connectivity, we integrate BLE 5.x, WiFi, LoRaWAN, Zigbee, Thread, LTE-M and NB-IoT — selecting the right technology based on range, data rate, power budget and deployment environment. We design LoRaWAN gateways, BLE-to-cloud bridges and MQTT brokers that serve as the backbone of connected product ecosystems. In industrial environments, we handle the translation between OT protocols and modern IT stacks, enabling legacy equipment to feed data into cloud systems without requiring infrastructure replacement. We also implement secure communication: TLS, certificate provisioning, mutual authentication and encrypted OTA firmware updates.

Cloud and backend systems designed specifically for the demands of connected hardware: high-frequency data ingestion, device management at scale, real-time telemetry processing and long-term time-series storage. We design and implement REST and MQTT APIs, WebSocket servers, message brokers and data pipelines that handle IoT data volumes reliably and cost-effectively. Infrastructure runs on AWS IoT Core, Google Cloud or Azure IoT Hub — whichever platform fits your existing ecosystem and compliance requirements. For time-series workloads, we select InfluxDB, TimescaleDB or managed cloud services based on query patterns, retention needs and cost. We build dashboards in Grafana or lightweight custom frontends that surface the right metrics to the right people. DevOps is built in from the start: services are containerised with Docker, infrastructure is managed as code with Terraform, and CI/CD pipelines with automated testing are configured from the first deployment. Security is an architectural constraint — TLS everywhere, certificate-based device authentication, role-based access control and secrets management are standard practice.

Complete, production-ready IoT products built by a single team from first principles — no integration gaps, no architectural inconsistencies between the device layer and the cloud, no finger-pointing between vendors when something does not work. We own the full technical vertical: hardware design, firmware, secure device-to-cloud connectivity, data ingestion, storage, analytics and operator-facing dashboards. This integration depth enables better decisions at every level: choosing the right communication protocol for the power budget, sizing cloud infrastructure for the actual data volume, designing the OTA update mechanism before the first prototype ships. A typical end-to-end IoT engagement covers requirements definition and system architecture, hardware prototype and firmware development, connectivity layer design — LoRaWAN, BLE, LTE-M or WiFi depending on the application — cloud backend with device registry and telemetry API, monitoring dashboard with alerting, and field deployment support including commissioning tooling.

Independent technical guidance for teams that need clarity on a complex engineering decision — without the bias of a vendor selling a particular platform or approach. We engage as a trusted technical partner to review existing system architectures, identify risks before they become expensive problems, and produce concrete, actionable recommendations covering embedded systems, firmware, hardware, connectivity and cloud architecture. Typical consulting engagements include: technology selection for a new product (MCU family, RTOS, cloud platform, communication protocol), architecture review before a significant development investment, hardware and firmware audit before a production run, technical due diligence for investors or acquirers evaluating a connected hardware company, and roadmap definition for products that need to scale beyond their initial design constraints. We are direct. We tell you what we see — the parts that are well-designed and the parts that carry risk — without padding reports with generic best practices.

Systematic, evidence-based validation for embedded systems and cloud backends — integrated into the development process from the architecture phase, not added at the end when problems are most expensive to fix. We define the test strategy early because catching a design flaw during schematic review costs a fraction of catching it in field-deployed hardware. For embedded systems, our QA practice covers unit and integration testing at the firmware level, hardware-in-the-loop (HIL) test rigs that simulate real sensor inputs and actuator loads under controlled conditions, functional test plans for board bring-up and production verification, power consumption profiling and endurance testing that exposes failure modes invisible in short-duration tests. For cloud and backend systems, we build automated regression suites, load testing pipelines, API contract tests and integration tests that run against real infrastructure. We also offer QA audits for teams with working products that lack confidence in their validation process: we review existing coverage, identify the highest-risk untested paths, and implement the infrastructure needed to give the engineering team a justified level of confidence.