Theme
Photo
PDF Download
Alexei Brown

ALEXEI BROWN

Software Engineer | AI, Sensor Fusion & Developer Tools

Sydney, Australia · contact via email · download PDF for full contact

qalarc.com/projects · linkedin.com/in/alexei-brown · github.com/fivelidz

Technical Skills

Languages
Python (primary) · JavaScript / TypeScript · SQL · Kotlin / Java (Android) · Familiarity with C / C++ and Go
AI / ML
PyTorch, TensorFlow, scikit-learn · pandas, NumPy, signal-processing pipelines · Deep learning, attention / transformer architectures · Computer vision (OpenCV, MediaPipe, ArcFace) · NLP, text classification, embeddings, LLM integration · Model evaluation, benchmarking, ablation studies · Behavioural intelligence over time-series sensor streams
Sensor & Signal
Software-Defined Radio — HackRF / Airspy / RTL-SDR panoramic scanning, RF spectrum analysis · Real-time DSP: decimation, phase-continuous VFO mixing, anti-alias filtering (rfai-dsp library, 18 tests) · Live waterfall/spectrum visualisation, demod pipelines (WFM/NFM/AM/SSB), RMS-based VAD · Bluetooth Low Energy (BLE) — scanning, fingerprinting, GATT exploration, protocol analysis · Wearable / phone sensor fusion (sleep, HR, HRV, accelerometer, rPPG) · Real-time inference on noisy, incomplete, adversarial data · Speech: Whisper STT (GPU-accelerated via faster-whisper/ROCm), voice activity detection
Infra & MLOps
Linux (Arch / CachyOS / NixOS) day-to-day · Docker / containerisation · Git, GitHub, automated CI/CD · OAuth 2.0 implementation · Cloud: Azure (Data Factory, Databricks), GCP (BigQuery), Cloudflare (Pages/Workers) · Edge / on-device inference: Ollama, ONNX Runtime Web, ROCm, quantised models
Robotics & HW
MediaPipe-based gesture / hand / face tracking pipelines · Servo control (FEETECH ST-3215-C047) on InMoov i2 robotic hand · SO-ARM101 robot arm with MuJoCo simulation + imitation learning · 3D printing for drones, cinewhoops, payload grabbers, perching legs

Experience

Developer Experience Engineer — Sahha · Remote (NZ-based company) · Current

Health Data Intelligence Platform — sensor data → behavioural intelligence

AI Engineer — Hypadrive · Auckland, NZ (remote) · Jun 2022 – 2024

Machine Learning & NLP

Legal Researcher (AI Ethics) — University of Newcastle Law School · Jun – Dec 2021

AI Regulation & Ethics

Neurotechnology Researcher — Collaborative Academic Projects · 2019 – 2023

AI applications in neuroscience

Selected Projects

Live demos and full write-ups on qalarc.com/projects

RFAI — RF Intelligence Platform qalarc.com/projects/rfai-monitor

SDR (HackRF / Airspy / RTL-SDR) → real-time demod → GPU Whisper STT → searchable archive — fully local

Multi-layer software-defined-radio intelligence stack: captures RF across UHF/VHF/HF/marine/aviation bands, demodulates in real time (WFM/NFM/AM/SSB), transcribes voice traffic with faster-whisper on AMD ROCm, extracts entities (callsigns, locations) and archives everything into a searchable SQLite/FTS5 database. The DSP core was extracted into a standalone library (rfai-dsp) with anti-alias decimation, phase-continuous VFO mixing and an abstract SDR backend.

Stack: Python · PyQt5 · HackRF / Airspy / RTL-SDR (SoapySDR) · DSP (rfai-dsp) · faster-whisper + ROCm · SQLite/FTS5

Smart Glasses Face Tracker qalarc.com/projects/smart-glasses-face-tracker · live demo

Wearable ISR: glasses → home computer (face recognition) → phone — custom ONNX demographic models

Real-time computer-vision pipeline for smart-glasses use cases: MediaPipe Face Mesh (468 landmarks), ArcFace recognition, multi-face tracking at 30fps with a persistent identity library — all in-browser. Ships custom-trained 'Folkus' demographic models (2.3 MB INT8 ONNX, running via ONNX Runtime Web) that beat commercial APIs: gender 92.1%, age 5.37y MAE, ethnicity 72.2% across 7 groups. Camera glasses stream to a host computer that recognises faces and feeds AI context back via phone.

Stack: JavaScript · MediaPipe · ArcFace · ONNX Runtime Web · Custom ONNX models · WebGL

Visual Biomarker Analyzer qalarc.com/projects/camera-biomarkers · live demo

Heart rate, HRV and SpO₂ from a webcam — 100% in-browser rPPG

Browser-based remote photoplethysmography pipeline extracting heart rate, HRV, SpO₂ and stress indicators from regular webcam video. No wearable required — pure signal processing on a noisy real-world input.

Stack: JavaScript · rPPG · Signal processing · Browser ML

chanalyse — Continuous OSINT + Auto-Journalism chanalyse.org

Real-time board scraper → NLP clustering → spike detection → auto-published sourced articles

Full-stack intelligence pipeline that continuously ingests public message boards, runs NLP for topic clustering, sentiment, authenticity and signal detection, flags narrative spikes via z-score detection, and auto-generates balanced, sourced journalistic articles with SEO/OG metadata. Live dashboard with rolling 24h/7d/30d windows; 5,400+ story artifacts and 14 auto-published articles to date.

Stack: Python · SQLite/FTS5 · NLP · LLM article generation · Cloudflare Pages

tradez.au — Directory + Website Builder tradez.au

Lead discovery → auto-generated mobile-first sites → deployed at scale on Cloudflare ($0 hosting)

End-to-end small-business platform: scrapes Google Maps for businesses with reviews but no website, qualifies them, auto-generates personalised mobile-first HTML sites (17 design variants) and deploys each to tradez.au/<slug>/ on Cloudflare Pages. 9,264 leads scraped, 1,657 qualified, 1,655 demo sites deployed, 4,568 businesses in the public directory. Includes CRM, outreach tooling and a 3-agent orchestration system.

Stack: Python · HTML/CSS · Cloudflare Pages/Workers · Web scraping · Stripe/PayID

Blue Truth — BLE Scanner qalarc.com/projects/blue-truth

Android BLE scanner with device fingerprinting, GATT explorer and Claude AI analysis

Advanced Android BLE scanner: device fingerprinting, GATT explorer, signal traces over time, behavioural pattern detection (proximity, dwell, repeat encounters), and built-in Claude AI analysis of captured BLE data.

Stack: Android · Kotlin / Java · BLE protocols · Time-series analysis

gmux — Gesture Terminal Multiplexer qalarc.com/projects/gmux · live demo

Gesture and voice layer on top of tmux for running multiple AI agents

Reimagining terminal multiplexing for AI-agent workflows: gesture-controlled agent dashboard, voice commands, tab cycling between tasks/chat/hardware, pinch-to-click overlay, 10 themes, live mock agent state.

Stack: TypeScript · MediaPipe · Terminal UI · WebGL

Orbital — N-Body Gravity Puzzle (Android) qalarc.com/projects/orbital

Native Kotlin + custom OpenGL ES 2.0 renderer, deterministic physics — 2.6 MB APK

Slingshot gravity puzzle built native in Kotlin (not Unity) for a 2.6 MB APK with instant cold-start. Deterministic semi-implicit Euler integrator with substeps (verified <1% drift over 20s for stable orbits), trajectory prediction that matches the live sim exactly, and a 4-pass OpenGL ES 2.0 renderer with procedural planet shaders, glow bloom and particle systems. 20 handcrafted levels + 3 sandbox systems.

Stack: Kotlin · OpenGL ES 2.0 (custom GLSL) · Android SDK 24–34 · Deterministic physics

Education

University of Technology Sydney

Master of Intellectual Property and Patent Law · In Progress

University of New South Wales

Bachelor of Science (Honours), Neuroscience & Pharmacology · First Class Honours

Generation Australia

Data Analytics Graduate Program

Publications & Presentations

Additional Experience

Teacher's Aide — University of Sydney & Newcastle Law School · 2013 – 2020

Co-Founder — Neuroscience Society, UNSW

Key Competencies

AI / ML Production ML pipelines, deep learning, NLP, behavioural intelligence, LLM integration
Sensor Fusion RF / SDR, BLE, wearables, rPPG, real-time inference on noisy real-world data
Computer Vision MediaPipe, ArcFace, OpenCV — multi-face tracking, gesture recognition, in-browser pipelines
Developer Tools CLI development, OAuth, cross-platform compatibility, terminal UX
Research Academic publishing, international presentations, ethical AI considerations
Communication Developer advocacy, technical writing, cross-functional collaboration