currently building in lagos, ng

I build systems
that run without me.

Autonomous bots, trading signal engines, and security tooling — designed once, deployed, and left running. If it can be scripted, scheduled, or automated, I've probably already built something like it.

Start a project → See what's live
live-systems.log
24/7
Uptime target
25+
Markets tracked
6
Signal strategies
100%
Python-built
$ ls --running

Systems in production

Each one started as a single script. Each one now runs unattended.

Apex Signals
A multi-strategy crypto signal engine watching 25 major pairs around the clock. Combines EMA crossovers, RSI, MACD, Bollinger Bands, volume confirmation, and live news sentiment into a single weighted score, then posts straight to Telegram with entry, take-profit, and stop-loss levels.
PythonTelegram APICoinGeckoTechnical Analysis
● LIVE
Nova — Personal AI Assistant
A JARVIS-style autonomous assistant running locally. Voice interface, laptop control, self-directed nightly learning, and a self-improvement loop that lets it read and extend its own code under supervision.
PythonLLM IntegrationSpeechAutomation
● LIVE
Recon Pipeline
An unattended security scanner that works through a target list automatically — subdomain enumeration, exposed secrets in client-side JS, CORS and header misconfigurations, reflected XSS and SQLi probing — and only pings a human when something is actually worth a look.
PythonReconAppSecTelegram Alerts
● ACTIVE
Contract Risk Scanner
Pulls verified source straight from Etherscan and flags the patterns that matter — stale compiler versions, missing reentrancy guards, tx.origin auth, unrestricted self-destruct — then scores the contract and writes a report.
PythonEtherscan APISolidity Analysis
● ACTIVE
$ cat approach.md

How a build actually goes

01 / SCOPE
Define the loop
Every automation is really a loop: trigger, decide, act. We nail down what starts it, what data it needs, and what "done" looks like before any code gets written.
02 / BUILD
Ship the working version first
A rough version running beats a polished spec sitting still. I build the core loop fast, get it actually running, then harden it — error handling, logging, retries.
03 / RUN
Leave it alone, on purpose
The goal is always unattended operation. If it needs me to babysit it, it isn't finished. Delivered systems come with the setup needed to run themselves.
$ pip list

Stack

Python python-telegram-bot Selenium BeautifulSoup Requests / HTTPX SQLite APScheduler REST APIs LLM Integration Web Scraping Bash / Shell Git
$ ./start_project.sh

Got something that should run on its own?

Telegram bots, scrapers, signal engines, scanners, scheduled jobs — describe the loop and I'll tell you exactly how it gets built.

Message on X → View on Contra