Analista incansable

Every day it watches a list of stocks, scores each company on its fundamentals and flags anything worth a second look. It analyses and watches; it never predicts or trades.

Personal project2026

Code on GitHub

Period: 2026
Team: Personal project
Context: Runs on my Ubuntu server, inside a private network
My role: The whole project
Tech stack: Python
pandas
yfinance
SQLite
FastAPI
Jinja2
n8n
Telegram

Watchlist sorted by distance to the entry zone, with each company’s score — Read-only web panel designed for the phone, with synthetic demo data.

Company detail with price, zone, RSI, moving averages and score breakdown — Read-only web panel designed for the phone, with synthetic demo data.

What it does

Downloads daily prices and annual statements into SQLite. Missing data is shown as “n/a”: it is never invented.
Scores each company from 0 to 100 on eight criteria and lists red flags separately.
Alerts when a stock enters its zone or gets within 3%, crosses the 200-day average or gets a new red flag.
Every zone alert includes the latest headlines and the manual note for the stock.

How it is built

Weights and thresholds in YAML, to tune the method without touching code.
An anti-spam table stops the same alert repeating for 5 days.
JSON output for an n8n workflow, or straight to the Telegram API.
A FastAPI panel with no write endpoints and a demo mode with synthetic data.

The score

Criterion	Green	Amber	Weight
Revenue growth	> 20 %	10 to 20%	20
Gross margin	> 50 %	35 to 50%	10
Free cash flow	positive and growing	positive	20
FCF margin	> 10 %	5 to 10%	10
Net debt / EBITDA	< 1x or net cash	1 to 3x	10
ROE	> 15 %	8 to 15%	10
Annual dilution	< 2 %	2 to 5%	10
Earnings keep up	grow with revenue	positive but slower	10

Green earns the full weight, amber half, and red or missing data zero. Red flags are listed separately and never averaged away.

Gallery

Daily report output in the terminal with zone, distance, RSI and moving averages for each stock — Daily report in the terminal, with synthetic data.

Score breakdown of a fictional company with its red flags — Score breakdown and red flags of a fictional company.

Alerts generated by the system with the manual note for each stock — Zone alerts ready to send on Telegram.

Let’s talk

I’m from Seville and now live in Madrid. I like things to work end to end: designing the system, writing the code, deploying it and measuring whether it really holds up.

If you’re looking for someone for backend, data or applied AI work, write to me and I’ll get back to you.