Gravelly real-time analytics interface used for AI-driven investment decisions

AI-led portfolio analysis, with your capital free to move

Gravelly applies predictive modelling to market data continuously, surfacing recommendations in plain language. Funds are never tied down by lock-up periods, so withdrawal requests are processed without the delays typical of traditional managed products.

The problem with manual analysis

Spreadsheets and static reports cannot keep pace with live markets

Most independent investors rely on information that is already out of date by the time it reaches a decision. Reviewing data manually takes time most people do not have, and by the time a pattern is spotted, the opportunity has often passed.

  • Analysis delayed by days, not minutes, when done by hand
  • Capital frequently locked away in funds with fixed notice periods
  • Technical dashboards built for analysts, not everyday investors

Gravelly was built to close that gap. Our models process incoming data continuously and translate the output into recommendations that do not require a background in statistics or coding to interpret.

See the advantages in detail
Gravelly data analysts reviewing predictive model output
How the engine works

Three mechanisms behind every recommendation

Each stage below is described in plain terms. You are not required to understand the mathematics; you are required to understand what the output means and why it was produced.

Predictive modelling

Anticipating market shifts before they occur

Rather than reacting to news after prices have already moved, Gravelly's models are trained on historical and live data to identify early indicators of change. The aim is not certainty — no model offers that — but an earlier, more informed read of likely direction, expressed as a confidence range rather than a single guess.

01 Historical patterns compared against current conditions
02 Live price and volume data layered on top
03 Output expressed as a probability range, not a prediction
Risk mitigation engine

Reducing exposure before losses compound

Every recommendation is paired with a risk assessment. The engine monitors volatility and correlation across holdings and will flag — or automatically adjust, depending on your settings — positions that move outside agreed tolerances. This does not remove risk; it is designed to keep it within boundaries you have set in advance.

01 Volatility thresholds set per holding
02 Continuous monitoring against those thresholds
03 Alert or rebalance triggered automatically
Real-time dashboard

One screen, updated continuously

The dashboard consolidates current positions, model confidence, and risk flags into a single view. Figures update as new data arrives, so you are looking at the same information the engine is acting on, not a summary produced hours earlier.

01 Positions and performance in one view
02 Model confidence shown alongside each signal
03 No separate reports to reconcile manually
Capital fluidity

Instant withdrawal, with no lock-up period attached

A common trade-off with managed products is that sophisticated analysis comes bundled with restricted access to your own money. Gravelly separates the two: the intelligence stays institutional-grade, but your capital remains yours to move.

1

Submit a withdrawal request

Request funds directly from your dashboard at any time, including outside normal business hours.

2

Position unwound automatically

The system closes the relevant positions without requiring manual intervention or a notice period.

3

Funds returned to your account

Cleared funds are returned to your linked UK bank account, with processing times shown before you confirm.

Processing times depend on underlying market liquidity and standard banking clearance windows; they are not instantaneous in the literal sense, but they are not subject to the fixed lock-up or redemption-window restrictions found in many managed funds. Capital is at risk, and withdrawal values reflect market conditions at the time of the request.

Methodology

What happens between data arriving and a recommendation appearing

We set out the process stage by stage below, rather than describing the model as an unexplained "black box". Each stage can be reviewed from your dashboard.

Stage 1

Data ingestion

Market feeds, pricing history, and volatility data are collected continuously and checked for completeness before being passed into the models. Incomplete or anomalous data is flagged and excluded rather than guessed at.

Stage 2

Neural optimisation

Pattern-recognition models weigh current conditions against historical analogues, adjusting their internal parameters as new outcomes confirm or contradict earlier assumptions. This is an ongoing recalibration, not a one-off training run.

Stage 3

Actionable output

Results are translated into a short, plain-language recommendation with a confidence range and risk rating, rather than raw statistical output. You decide whether to act, adjust, or decline the suggestion.

Frequently asked

Questions we are asked most often

If your question is not answered here, our contact page or the full FAQ page covers further detail.

Technical barriers

Do I need any coding or data analysis experience to use Gravelly?

No. Recommendations are presented in plain English with a confidence range and a short explanation of the reasoning behind them. The underlying models require no configuration from you.

How much time does Gravelly require from me each week?

Most users check the dashboard a few times a week. The system does not require daily attention to function, though you can review activity as often as you wish.

Withdrawal mechanics

Is there a minimum holding period before I can withdraw?

No. There is no fixed lock-up period. Withdrawal requests are processed as soon as the relevant positions can be unwound, subject to standard market and banking clearance times.

Will withdrawing early reduce the value I receive?

Your withdrawal reflects the market value of your holdings at the time the request is processed. There is no early-exit penalty charged by Gravelly itself.

Security protocols

How is my account access protected?

Account access uses standard encrypted connections and authentication checks at login. Full details of our security practices are available on request via the contact page.

Can I see exactly why a recommendation was made?

Yes. Every recommendation includes a short rationale, the confidence range used, and the risk rating applied, so you are not acting on an unexplained output.

Review the current analysis before you decide anything further

Create an account to view live model output on your own dashboard. There is no obligation to commit capital at this stage.

Start Optimising No credit card required to view your first analysis.