“The project works well with the market and helps build a flexible strategy for futures trading in the crypto market. As for the downsides, I would mention the interface — at first, it does not seem very clear, so it takes some time to figure it out. But overall, the project left a positive impression.”
Quant lab in the cloud · self-hosted execution
Is your edge real?
See for yourself.
Fast to
Write the strategy once, as explicit rules. The same definition is backtested with real costs, checked on unseen data and run live. One evening ends with a decision: run, rework or drop it.
7 days free, no card
Research to live
1
engine for the backtest and the live run: what you validated is what runs
The Execution Log puts every live fill, including stops and take-profits, next to what the test expected. You see where live differs. On trades up to $300, entry price, slippage and costs land within 0.5% of the test.
Validation
4
checks before you run a strategy: unseen data, random trade order, market conditions, other assets
Part of the history stays hidden from the optimizer and tests the result afterwards. The same trades in a random order show the possible drawdowns. Each check gives its own answer, and the decision stays with you.
Portfolio
30
strategies run at the same time from one account, each on the exchange you pick
Trader runs 5 live strategies, Pro runs 30. Set the risk once: the size and leverage of every trade are computed for you, with the liquidation price behind the stop. No need to sit at the chart: you review the trades in the journal when it suits you.
AI agent
Betelgeuse
AI agent that researches strategies with you in a chat
It is built on the agent we use in our own research. Ask it to build a strategy, test an idea or explain a result. It points out the weak spot and stops when the evidence does not hold up. Included in Trader and Pro; on Pro it also draws on our research knowledge base.
An entry is not a strategy.
It is only the first step. Seven more decide how the trade is managed:
EntryStop-lossTake-profitsTrailing stopEarly exitBreak-evenSkip tradesDynamic risk
Any of them can be driven by an indicator.
And the work does not end at launch:
ResearchTradeLearnImprove
You trade the exact strategy you tested. The Execution Log shows every live fill next to what the test expected. What you learn goes into the next version of the rules.
From idea to a decision in 4 steps
1
Build the strategy from rules
Pick indicators, set the entry and exit rules, add a stop-loss and take-profits. Every rule stays readable.
2
See what it makes after real costs
The backtest charges your exchange fee, funding and slippage. You see every trade, 20 metrics and the drawdowns you could face.
3
Find parameters that hold up
The optimizer tries thousands of combinations and tests the winner on history it never saw. A strategy that only fits the past fails here, before it reaches your account.
4
Run the same strategy live
Assistant mode sends you the signal and you place the order. Auto mode runs an open-source app on your machine around the clock. Both run the definition you tested, and the Execution Log shows every live fill next to what the test expected.
Pick indicators, set the entry and exit rules, add a stop-loss and take-profits. Every rule stays readable.
The backtest charges your exchange fee, funding and slippage. You see every trade, 20 metrics and the drawdowns you could face.
The optimizer tries thousands of combinations and tests the winner on history it never saw. A strategy that only fits the past fails here, before it reaches your account.
Assistant mode sends you the signal and you place the order. Auto mode runs an open-source app on your machine around the clock. Both run the definition you tested, and the Execution Log shows every live fill next to what the test expected.
Two decisions that shape everything else.
Both were made before the product existed, and every feature since then has had to fit them.
See why every trade was taken or skipped
If a number surprises you, you have to be able to find out why it came out that way.
Skipped entries come with a reason: blocked by your own rules, or refused by the engine.
Your signals, your executor
The signal is yours over the socket, execution is yours, and the rules stay yours as readable conditions that work outside our platform.
Between the signal and the exchange you can put your own filter, model or executor.
You get the test, the run on unseen data and the Monte Carlo spread. You make the decision.
Exchange connections
Each strategy picks its own exchange, and several can run side by side from one account. Market data comes over the exchange API.
Supported
Supported
Supported
Supported
Soon
COINBASE
Soon
KuCoin × Veskald
Veskald is an official KuCoin broker partner. Open your KuCoin account with referral code VESKALD or our link. We receive a share of the exchange fees, and it costs you nothing.
One definition instead of four implementations.
A strategy is usually rebuilt for every tool: Pine for the idea, Python for the backtest, a bot config for live, a spreadsheet for the journal. The versions drift apart. Here the backtester, the optimizer, the bot and the journal read one definition. Change it in one place, and every tool sees the change.
TradingViewIdea
Typical stackVeskald
Backtest on real costs$
Optimizer$
Self-hosted bot, your keys$
Trade journal & Execution Log$
Strategy portfolio—
AI agent—
TradingView
Idea
Typical stack
a separate implementation each
$Backtest framework
$Optimizer
$Algo bot service
$Trading journal
Start free. Pay when you see what works.
The Free plan gives you 7 days to build and test your first strategy. Paid plans are for when you are ready to run several strategies in parallel and manage a portfolio.
Multi-timeframe entry conditions, stop-loss, multiple take-profits, trailing stop, early exit, move to break-even, dynamic risk management and trade-skip rules. Progress saves at every step.
On perps, funding alone can eat the whole profit — here it is part of every simulated trade. Side-by-side comparison and an Out-of-Sample period separate real edge from a lucky streak.
The stress test shows how robust the strategy is and the probability of losses in live trading.
Breakdown by volatility, trend, volume, momentum and trading session: the exact conditions where the strategy earns and where it loses, with concrete recommendations for improvement.
Sweeps combinations of entries, stops, take-profits, break-even moves, early exits and skip rules. The out-of-sample split keeps the strategy from memorizing history.
Technical indicators and their complex variations (divergences, slopes); AI reads the news and aligns it with the chart by release time; liquidity, the dollar index, inflation and the U.S. economy; an extended order book with filtering.
Risk is configured per strategy; the potential loss is known before you enter the trade.
Days of the week and hours are set for each strategy individually.
Ask Betelgeuse to build a strategy, test an idea or explain a result. It points out the weak spot and stops when the evidence does not hold up. Comes with Trader and Pro, with a free monthly minimum of tokens. On Pro it also works with our research knowledge base.
Assistant: read-only keys, prompts via Telegram, email or browser — when to move a stop, take profit or react to the news. Automatic: an open-source app on your own computer (Windows / macOS / Linux) or a server, where a single command does the whole setup (Docker, from $4/mo). Any difficulty or question — we’ll help.
How the backtester ran the trade, what notifications were sent, and what was actually executed. If you deviated from the plan, you see whether it helped or hurt you.
On each candle: what happened, the indicator values and the news released at that moment. You see why a trade worked or failed.
The strategy is re-run across other coins to separate a working rule set from a coincidence on a single chart.
Combined equity curve, correlations between strategies, drawdown compensation — and each strategy’s real execution tracked individually.
Optimization runs on a dedicated server at full speed; setup and configuration take up to three business days.
Developed to your specification; available only to your account.
Your subscription grants access to software tools. Veskald does not provide investment services and does not guarantee trading results.
Frequently asked questions
Who is Veskald for?
Systematic crypto traders who already research their own strategies. Veskald gives them a shorter path from an idea to evidence they can check. The strategy they tested is the one that runs live. They skip building their own research tools.
Will my live results match the backtest?
Close, but not to the cent. The live run uses the same rules and price data, and the test already charges fees, funding and slippage. On trades up to $300, entry price, slippage and costs land within 0.5% of the test. The Execution Log shows each difference.
Do I need to know how to code?
You build strategies as explicit rules, so no strategy code is needed. Pick indicators and conditions, combine them with AND/OR and add filters from other timeframes. If an indicator is missing, ask for it and we build it for everyone.
Can I start from a ready strategy?
Yes. The library holds strategies we research and publish ourselves, with the logic open. Pick one, see its tests, then change it or run it as is. You work on a copy, so the original stays intact.
Can I check a strategy I bought somewhere?
Yes, if you know its rules. Rebuild them in the builder, then run the full check: real costs, unseen data, Monte Carlo and other assets.
How is my exchange account protected?
Veskald asks only for a read-only key, which lets it see positions and balances. Orders are placed by the open-source app on your own machine, with trading keys that stay there. The full setup is on the Security page.
What happens after the 7 free days?
When a strategy is ready to go live, switch to Trader or Pro. The 7 days include 3 strategies, the backtest with Monte Carlo and cluster analysis, and 50 optimizer combinations. You also get the market report. No card needed. The Veskald AI agent comes with Trader and Pro.
Which exchanges are supported?
Stable today: Binance, Bybit, Kraken, KuCoin, Hyperliquid. In development: OKX, Coinbase. If you need another exchange (Bitget, BingX, dYdX and others), write to us. We usually add it in 1-3 weeks.
How is the optimizer different from my manual grid search?
Manual search covers what you have time for, 10-20 combinations. The algorithm tries thousands and ranks them by how smooth the equity curve is, so one lucky trade cannot win. It searches on one part of the history, and the part it never saw tests the result afterwards.
Can I add my own indicator?
Ask for it and we build it into the platform for everyone, so every strategy stays readable and testable. The library already has 100+ indicators and candlestick patterns, from RSI to Smart Money Concepts. Your own code can run after the signal, in the open-source bot or your own executor.
Can I connect Veskald to my own systems?
Yes, on Trader and above. Each strategy streams its events over a WebSocket to an endpoint you register. Put your own filter or model before the order, or replace the open-source bot with your own executor. Strategy files have no one-click export, and the rules stay readable conditions you can rebuild anywhere.
How often do updates ship?
Releases ship 1-2 times a month. Active users shape the roadmap, and requests from tickets go into the work. The latest major release is the Veskald AI agent, Betelgeuse, which researches strategies with you.
Who can read my strategies?
With two-key encryption on, only you. Strategies are stored encrypted in the Veskald database, where backtests and the optimizer run. Turn on two-key encryption, and nobody opens a strategy without your key, the Veskald team included.
Where is my data stored?
On servers in the EU (Hetzner, Germany). GDPR applies in full, details are in the Privacy Policy. Strategy encryption with a two-key scheme is available for your trading logic.
Do you use my strategies or backtests to train your AI or build your own strategies?
Your strategies are used only to run the service for you: backtests, optimization and launch. The public template library comes from our own research agent. On Pro, Betelgeuse reads that knowledge base with you and writes nothing back. With two-key encryption on, nobody reads a strategy without your key, us included.
Reviews on Trustpilot
“The platform is new but already offers features that are hard to find on similar resources. + Everything in one place, a trial, flexible customization and AI. - It’s a young startup. I liked it and have started using it.”
“Nie przypuszczałem, że kiedykolwiek w życiu będę mógł skalować swoje pomysły z jednego miejsca na różne platformy tradingowe i mieć przy tym tak wysoki poziom kontroli. Polecam!”
Build your first strategy: 7 days free, no card
Write the rules, test them on real data and end the evening with a decision: run, rework or drop it.
Risk Disclosure
Trading crypto-asset derivatives, including perpetual futures, involves substantial risk of loss and is not suitable for every market participant. Veskald provides software tools for building and testing trading strategies and for executing them through an app on the user’s own machine. This is not investment advice, not asset management, and not a guarantee of trading results. Past backtest performance does not predict future returns.