Where time
meets intelligence.
I'm Michael-David Osuji. I build AI-driven automation systems, trading infrastructure, and procurement intelligence platforms that run without supervision, from first contact to final report.
Automation that replaces the manual grind, not just the busywork.
I'm a Python automation and AI engineer working out of Lagos. My work sits at the intersection of LLM orchestration, browser and API automation, and production backend engineering: taking workflows that currently depend on a person clicking through a login-restricted platform, and turning them into systems that run on a schedule and report back.
I hold a BS in Computer Science from the University of the People, and I currently work as a Python Automation Engineer under contract with a US-based consulting firm, alongside direct freelance client work.
Recent client work has included reverse engineering login-restricted government procurement platforms with no public API, and packaging the resulting automation as standalone Windows executables for non-technical end users.
Stack
Systems in production.
Status reflects the current state of each system: deployed and running, actively live in a market, or still in development.
A full-stack SaaS platform that automates interaction with the U.S. Defense Logistics Agency's DIBBS procurement system: session-based auth, anti-bot navigation, PDF ingestion, and regex-driven extraction of NAICS codes, set-aside status, and pricing history.
A deep-learning forecasting system built around a Temporal Convolutional Network, engineering 70+ statistical and technical features across 550+ crypto assets for multi-target High/Low/Close prediction.
An automated Binance Futures trading system using multi-timeframe signal architecture (1H/15M). Google Gemini analyzes processed market data and returns structured trade decisions: direction, entry, stop-loss, take-profit, and confidence.
A production-style SaaS backend for document automation: ingests invoices and contracts, runs OCR, classifies document type, and extracts structured fields through an async FastAPI + Celery pipeline.
A workflow automation backend where teams define step-based automations in JSON and run them manually or on a schedule: scraping, transforming, and exporting data with full run history.
A real-time computer vision system recognizing standardized crane hand signals from live video, built for construction safety monitoring and as a second set of eyes for crane operators.
A plant disease classification system with ten species-specific CNN models, covering 3 to 10 diseases per species, built to support early intervention and reduce unnecessary chemical treatment.
Single-purpose web tools.
Free to try in the browser, no signup. Paid tiers unlock through a Gumroad license key, right on the page.
Removes em-dashes and rewrites the punctuation around them properly, semicolon, colon, comma, or a clean new sentence, instead of just deleting them.
Invoice and quote generator with PDF export, for anyone who needs a clean, professional document out in minutes, freelancers, contractors, consultants, and small businesses alike.
Where I plug in.
AI Automation Systems
LLM-powered pipelines that read, decide, and act: signal generation, document processing, and structured decision-making from unstructured input.
Trading Infrastructure
Multi-timeframe signal architectures, risk-managed execution engines, and forecasting models for crypto and equity markets.
Procurement & Business Intelligence
Automation for login-restricted platforms with no public API: session handling, structured extraction, and reporting for government and B2B data.
Backend & Data Pipelines
Production FastAPI/Flask backends, async processing with Celery and Redis, and clean data extraction and organization for analysis or training.
Tell me what needs to run on its own.
Send a message directly, or use the form. Everything submitted here goes straight to my inbox.