Building in public · Available for work

Hugo Kuznicki

Building AI automation & trading tools in public.

I design and ship practical automation systems — local LLM apps, market scanners, and data pipelines that run fast, run private, and earn their keep. Real tools, shipped end-to-end, documented in the open.

  • 4+shipped projects
  • 100%local-first builds
  • Python · JScore stack
// selected work

Projects, shipped

Each of these runs end-to-end. Built solo, documented in public.

Market Scanner dashboard screenshot [SCREENSHOT]

Market Scanner

A fast, local market screener that pulls daily OHLCV data, computes technical signals, and surfaces opportunities in a sortable dashboard with sparklines and per-ticker charts.

  • Python
  • Flask
  • yfinance
  • pandas
  • vanilla JS
Content Studio UI screenshot [SCREENSHOT]

Content Studio

A local web app that drafts LinkedIn and X posts on demand — streamed live from your own Ollama models. No cloud, no API keys, zero cost per post.

  • Python
  • FastAPI
  • Ollama
  • SSE
  • local LLM
[SCREENSHOT]

Content Dashboard

A production board for a faceless short-form video channel — tracks every video from script → clips → voiceover → assembled → posted, with one-click copy into ElevenLabs, Kling, and CapCut.

  • HTML
  • CSS
  • JavaScript
  • workflow
Review Miner report screenshot [SCREENSHOT]

Review Miner

Paste a pile of customer reviews and get back a structured report — recurring themes, praise, pain points, and the top asks — so you know exactly what to build or fix next.

  • Python
  • local LLM
  • NLP
  • reporting
Listing SEO Generator dashboard screenshot [SCREENSHOT]

Listing SEO Generator

An SEO copy generator for Etsy & Shopify sellers — turns a product into optimized titles, tags, and descriptions in one pass. Batch-ready and export-friendly.

  • Python
  • local LLM
  • SEO
  • e-commerce
[CASE STUDY]

Apex Wallet

🔒

A rules-governed trading wallet: it ingests signals, validates each against a hard-coded risk doctrine, and trades — paper by default, live only as a deliberate, multi-step action. The rules engine is a seatbelt for aggressive strategies.

  • Python
  • risk engine
  • ccxt
  • backtesting
Private · Case study Ask for a walkthrough →
// about

AI-automation builder, working in the open

I'm Hugo — I build AI-automation and trading/data systems, end to end. My focus is practical software that does real work: local-first LLM apps, market scanners, and the data pipelines that feed them.

I prefer tools that run on your own machine — fast, private, and free of per-call API costs. I ship in public, document as I go, and care about clean interfaces as much as the engine underneath.

Right now I'm available for AI-automation work. If you have a workflow that's begging to be automated, or a data problem that needs a sharp tool, let's talk.

// work with me

How I can help

Fixed-scope AI-automation engagements. Pricing below is placeholder — easy to edit.

Automation Sprint

A single painful workflow, automated end-to-end — scoped, built, and handed over with docs.

  • Discovery + scope call
  • One working automation
  • Docs & handover
From $X,XXX · fixed price
Start a sprint
Most popular
🤖

Local LLM App

A custom local-first AI app — content generation, internal copilots, document tools — running on your own models.

  • Custom UI + API
  • Runs on local Ollama / your stack
  • Zero per-call API cost
From $X,XXX · fixed price
Build my app
📈

Trading / Data Tooling

Market scanners, signal pipelines, dashboards, and risk tooling — built to your strategy and risk rules.

  • Data ingestion + signals
  • Dashboard or CLI
  • Paper-first, rules-governed
From $X,XXX · fixed price
Discuss a tool

Not sure which fits? Tell me what you're trying to automate and I'll point you at the right scope.