Hey Product Hunt!

I'm Eugene, one of the makers of Veskald — and here's the honest story of how we got here.

We're developers. For years we built products — some of us freelancing, some at bigger companies. At some point the work turned into routine, and we wanted something new. Trading caught our attention, and crypto looked like the lower-barrier way in: register, click, trade. Easy.

Getting in wasn't the problem. The problem was the wall of crypto coaches waiting on the other side. The deeper we dug, the more it looked like nothing they were selling actually worked (maybe we got unlucky, maybe that's just how it is). You can't buy every course in the world — and you don't need to. Trading was the horse we wanted to break in, and we wanted to understand how it actually works underneath the promises and the PnL screenshots.

So we ignored the noise, picked up a stack of books, and started practicing. Two things became obvious:

Trading is genuinely hard work — concentration, time, attention, discipline, self-control.
Most of it can be automated.

2022 — the first attempts

We tried the well-known bots on the market. The functionality always disappointed, and there was this lingering unease — too much black box, not enough control. Something always felt missing.

So we decided to build our own, from scratch, for personal use. We had enough engineering experience between us. The first thing we laid down was the architecture, and the rule was simple: every component independent, every component responsible for exactly one job. That way the system would be easy to adapt later.

We knew what we needed:

A solid backtester
An optimizer that nudges parameters while trying to avoid overfitting
An execution layer
Two modes: fully automated, and manual with the system as your assistant
Position sizing that actually respects your risk — because in trading, losses and profits walk hand in hand. That's just the job.

Six months in, the project looked completely insane. Eight or nine console apps running in parallel. Any change to a strategy, any test of a new idea meant writing code and restarting everything. It was slowing us down.

The pivot to a real product

We needed an interface. We also needed to move the whole thing onto servers — locally, it had become unmanageable. So we built the UI and wired everything together.

The result:

Technical analysis according to your settings
Notifications when the setup matches, with position size pre-calculated so the loss is always controlled
Trade tracking with alerts that tell you what to do and when

All of it inside a visual builder where you configure multi-timeframe logic, indicators per timeframe, stop-losses, trailing stops, and breakeven moves — no code.

Useful, but still missing two layers: an analytics layer to help you make better decisions in semi-auto mode, and an execution layer for full automation. So we built both. That's how the Analytics module and the open-source execution bot were born.

The pain of manual trading

In manual mode, one thing kept bothering me. You'd miss an important signal on some timeframe because you were tired, or because the chart had turned into noise. I wanted the whole picture on one screen — already processed.

So we built it: candle patterns, divergences, support/resistance, and my favorite feature — signal accumulation. After each candle closes, the system marks what happened on it: red for negative, green for positive, yellow for mixed, gray for neutral. Seeing every timeframe at once, marked this way, changed how we traded.

That led to the next step. If we have an open trade and negative factors start stacking up on the chart, the system now warns us. The open-source bot does its job in parallel: connect to the socket, wait for commands, execute. Simple by design.

Still wasn't enough. We sat down, looked at the whole system, and wrote down what was still missing:

An auto-journal with candle-by-candle replay. For every candle: what I did (moved a stop, took partial profit), what news came out at that moment, what the TA was showing.
An Execution Log — because in manual trading we kept deviating from the plan. Three columns: how the trade actually went, how it should have gone per the backtest, and what signals we received in real time.
Portfolio analysis — we knew how one strategy behaved, but what happens with 2, 3, 10, 20 running together? How do they correlate? When do they bleed at the same time?
Strategy monitoring — see who's earning and who's losing, so we can pause or rework the bad ones.
More backtest metrics, plus Monte Carlo permutation — we were tired of analyzing every trade by hand. We wanted a rough map of what the future might look like.

And answers to all the "what if" questions that never stop coming.

Where we landed

Veskald is now a builder of trading assistants and bots, with:

Strategy builder — multi-timeframe, stop-losses, take-profits, trailing stops, breakeven, early exit, dynamic risk management. Every piece configurable, every piece optional.
Backtester — tunable core that accounts for slippage, commissions, and funding. Beyond the basic metrics: Monte Carlo, and automatic grouping of trades by market regime (volatility, trend, volume, momentum) and session.
Execution — an autonomous, open-source script you run locally or on a cheap VPS. Nobody shares trading keys with us.
Manual mode — analytics and alerts walk you through the trade.
Portfolio — combine strategies, see correlations, see where they lose together.
Journal — every angle of every trade.
Risk management — position size calculated automatically, loss boundaries respected.
Time management — when to trade, when to be notified.

Why we're shipping it

You'd think with all this, we'd be set. But the infrastructure has grown to a scale where maintaining it eats almost all of our time — there's barely any left for chasing new ideas. And I haven't even mentioned the AI we use to tag news articles; it needs constant retraining. Exchange APIs change. The "what if" questions never end.

So we made a call: this should belong to other people too. That's how Veskald became a product. There's a free plan, no card required — build a strategy, run a full backtest with Monte Carlo, and decide afterwards.

Happy to answer anything in the comments — especially honest critique.

Thank you for taking the time.

Eugene