AI-powered investment insights
We make it possible to follow, understand and copy decisions from AI strategies that continuously analyze market data. You choose the direction, the platform executes the trades.
The challenge
Most first-time investors encounter a stream of numbers, news and recommendations that rarely point in the same direction. It's not a lack of information that's the problem — it's the amount of it.
Pivuro Lemalu is based on predictive models, that is, statistical models that analyze historical and current data patterns to assess likely developments. They filter the noise from market data so that the individual investor sees a smaller but more relevant slice of information. It's the same type of analysis institutional investors have had access to for decades, made available in a simpler format.
This is how it works
The process is designed to be transparent. You don't need to understand the math of the models to understand what they do.
Step 1
The models continuously review millions of data points — price movements, trading volume, news flow and macroeconomic indicators. The purpose is to identify patterns that are difficult to see with the naked eye.
Step 2
You choose a strategy that matches your risk profile and time horizon. Each strategy is built on a different analytical logic, described in plain language, so that the choice is informed rather than guesswork.
Step 3
When a strategy generates a signal, the platform executes the trade according to the rules you approved in advance. You can always see what has been done and why.
Strategies
We describe the methodology behind each strategy because we believe that understanding the logic is more important than a number on a screen.
Focuses on limiting fluctuations by responding to changes in market volatility, that is, how much prices move over time. The model reduces exposure when uncertainty increases.
Identifies persistent price movements based on historical patterns and trading volume. The strategy takes a more active position when several data points point in the same direction over time.
Assesses the tone of news and public communications using language models, and compares it to price movements. The strategy continuously adjusts as the underlying mood changes.
Method and transparency
Pivuro Engine is the analytical core of Pivuro Lemalu. It combines multiple data sources — price data, volume, news flow and macroeconomic indicators — to avoid a single source or assumption being given too much weight in a decision.
The models are continually reassessed as new data comes in. This means that a strategy can adjust its position if the underlying assumptions change, without waiting for a monthly review.
We have deliberately refrained from building a model that tries to predict everything. Diversified data inputs mean that no single source of error can dominate the decision, but it also means that no model can guarantee an outcome.
Investing always involves the risk of loss, no matter how sophisticated the underlying analysis. AI strategies are decision support based on historical and current data — they are not a prediction of future results. We recommend that you only invest funds that you can bear to see fall in value and that you choose a risk profile that reflects your time horizon.
Frequently asked questions
Your funds are held at a regulated custodian bank, and the platform does not have direct access to withdraw funds from your account. However, this does not change the fact that investing in itself entails a risk of loss. For us, security is about two things: protecting your data and funds, and transparency about how decisions are made, so you're never in doubt about what's behind a trade.
There is a transparent fee attached to following a given strategy, which is shown before you confirm your choice. We do not charge hidden trading fees beyond what is described in your agreement. You can change strategy at any time or stop completely without commitment.
The models re-evaluate their assumptions continuously as new data comes in — this is called incremental learning. This means that a strategy can adjust its weighting of data sources over time, but it does not change its basic methodology without human review. A team follows the behavior of the models and intervenes if data or market conditions deviate significantly from what the model was built to handle.