Nokia • Telecom RAN Reliability • Oulu, Finland
Embedded / System Engineer
Cross-layer HW/SW Debug • Reliability Engineering • AI-assisted Diagnostics
Independent R&D: Edge AI • Embedded Diagnostics • Local Inference
Embedded/system engineer with 9+ years in electronics manufacturing and 3+ years in embedded development. My professional work focuses on system-level HW/SW troubleshooting, reliability engineering, Linux-based diagnostics, automation, and evidence-driven root-cause analysis in telecom environments.
I build tools and workflows for parsing large diagnostic datasets, correlating signals, and turning ambiguous system behavior into reproducible engineering hypotheses. AI is used as an engineering assistant for triage, pattern discovery, summarization, and tool orchestration — while raw technical evidence remains the source of truth.
Outside my employment, I explore independent R&D in local AI, embedded diagnostics, reliability tooling, and edge-computing workflows using personally owned equipment together with public, open-source, or synthetic datasets.
My hands-on background also includes schematic design, PCB development, component selection, bring-up, embedded Linux, RTOS development, and low-level interfaces including I2C, SPI, UART, SDIO, PCIe, BLE, Zigbee, and Z-Wave.
Systematic investigation across hardware, firmware, Linux, storage, interfaces, and environmental factors. From symptom collection and hypothesis generation through controlled validation and root-cause isolation.
Python-based diagnostic pipelines for parsing technical data, extracting and correlating signals, and generating consistent engineering outputs. Local LLM and agent-based tools for triage, hypothesis generation, retrieval, and workflow automation — while deterministic evidence remains authoritative.
From schematic capture through PCB layout to hardware bring-up. Embedded Linux drivers, firmware, and low-level interface integration. IoT connectivity with BLE/Zephyr, Zigbee, Z-Wave.
My independent work is developed outside my employment using personally owned hardware and software environments.
Projects in this section are based on:
No employer source code, internal documentation, customer/network data, proprietary datasets, internal tools, or confidential technical material is used in these projects.
Independent investigation framework for storage and filesystem reliability in embedded Linux systems. Focus areas include unexpected power interruption, filesystem integrity, write behavior, telemetry collection, reproducible fault injection, and evidence-based attribution of storage-related failures. Environment: personally owned hardware, public documentation, open-source tooling, and synthetic/reproducible test workloads.
Experimental methods for identifying intermittent board-level and interconnect faults from system telemetry and controlled stress testing. The project explores correlations between resets, environmental variables, error counters, and repeatable electrical or thermal stimulus. Environment: generic embedded test hardware and synthetic fault scenarios.
Hybrid diagnostic framework combining deterministic parsers with local AI agents for unfamiliar patterns and multi-source evidence correlation. Pipeline: raw logs/telemetry → deterministic parsing → feature extraction → retrieval/context → specialist agents → evidence correlation → human-reviewed RCA report. Development uses public, open-source, self-generated, and synthetic engineering datasets.
IoT/edge gateway for protocol bridging, sensor aggregation, local processing, and smart-environment monitoring. Designed as a personal embedded-systems test platform for connectivity and edge-data experiments.
nRF52-based sensor nodes with e-ink display, Zephyr RTOS firmware. BME680/SHTC3 environmental sensing with motion detection. CR2032 battery-focused power optimization. Also used as a controllable source of environmental telemetry for independent edge-AI experiments.
Local Diagnostic Intelligence
Masterbear explores hybrid diagnostic architectures that combine deterministic engineering tools with local language models and agentic workflows. The goal is to correlate logs, hardware health, storage behavior, environmental telemetry, and system events — while preserving traceable technical evidence. Local-first, privacy-preserving, evidence-driven.
Local inference for engineering workflows
Compound / multi-agent diagnostic systems
Retrieval over technical documentation
Telemetry correlation
Synthetic fault generation
Storage, power, and thermal reliability
Human-in-the-loop root-cause analysis
Local-first
Evidence-driven
Reproducible
Hardware-aware
Privacy-preserving
Open to discussions on embedded systems, reliability engineering, local AI, diagnostic automation, edge computing, and AI-assisted engineering workflows. Professional telecom experience and independent R&D projects are presented separately.