Vermeil Rentelle — predictive analytics platform for investors

Predictive analytics for better informed investment decisions

Vermeil Rentelle continuously processes financial and operational data streams to produce quantified recommendations. Each signal is time-stamped and viewable in a public log, with no promise of gain, only measured results.

Optimize my strategy Access performance logs

Dashboard overview: consolidated data feeds, confidence scores by asset and risk alerts updated in real time.

Our approach

A method designed to reduce uncertainty, not mask it

Vermeil Rentelle combines statistical models and machine learning to transform heterogeneous data volumes into actionable indicators. The goal is not to predict with absolute certainty, but to provide a documented decision-making basis, continually revised as new data arrives.

The platform is aimed at professional investors and project leaders who manage their activity alongside a main job and are looking for reliable analytical support, without having to constantly monitor the markets.

Vermeil Rentelle — technical team working on data analysis models
Abilities

Three technical pillars to support the decision

Each module operates independently but feeds the same data base, so that the recommendations remain consistent with each other.

01 — Modeling

Predictive analytics applied to market trends

The models are trained on historical series and recalibrated at regular intervals. They produce probabilistic scenarios rather than a single value, to represent real uncertainty rather than hiding it behind a number.

02 — Real time

Continuous analysis of data streams

New market, economic and sectoral data are integrated as they are published. The indicators displayed reflect the latest update available, with a timestamp visible for each signal.

03 — Risk

Structured risk reduction

Each recommendation is accompanied by a confidence index and an expected range of variation. The user then decides on the allocation or action, Vermeil Rentelle does not replace final human judgment.

Methodology

How a recommendation is produced

The technical logic is documented so that each user understands the origin of a signal before exploiting it.

Step 1

Data ingestion and standardization

Market sources, macroeconomic indicators and sectoral data are collected and then standardized in a common format, with consistency checks on entry.

Step 2

Continuous model calibration

The models are re-evaluated against the actual observed results. The differences observed are used to adjust the weightings before producing the next signal.

Step 3

Return of recommendations

The result is presented in the form of a score, confidence range and summary justification, viewable from the dashboard or exportable for internal use.

Transparency

Community Verified Performance Log

Past signals remain viewable and are never removed from the log, even when the observed result deviates from the forecast.

Journal audited by the user community
Period Sector analyzed Published signals Directional accuracy observed
T1 Large cap stocks 184 Illustrative excerpt
T2 Currencies and rates 92 Illustrative excerpt
T3 Raw materials 67 Illustrative excerpt

The values above illustrate the log structure and are not guaranteed results. The real figures, updated continuously, can be consulted after creating an account and vary depending on the period and sector observed. Past performance is no guarantee of future results.

Use cases

Three profiles, three distinct uses of the same data

The platform adapts the presentation of the signals according to the desired objective, without modifying the underlying calculation method.

Institutional investors

Multi-asset exposure tracking

Consolidation of positions and risk signals across multiple asset classes, with data export for investment committees.

Strategic planning

Business decision support

Analysis of sectoral scenarios to guide budget allocation or development choices, without replacing internal expertise.

Portfolio management

Risk-adjusted return optimization

Periodic adjustment of the distribution of a portfolio according to the confidence scores produced by the models, to be reviewed manually before execution.

Technical questions

Security, integration and subscription operation

The following points cover the most common questions asked before going live.

How is data protected?

Transmitted data is encrypted in transit and at rest. Account credentials are never shared with third parties and access to detailed logs requires dedicated authentication.

Can the platform integrate with our existing tools?

Recommendations are accessible via structured export, which allows integration into an internal dashboard or portfolio management tool already in place. Full API integration depends on subscription level.

How does the subscription work?

Access is offered in the form of a recurring subscription, with no minimum duration commitment imposed beyond the current billing period. Details of access levels are presented before registration.

Go from intuition to a documented decision

Consult the performance log before registering, then activate the recommendations corresponding to your risk profile.

Access performance logs

No credit card required to consult the public newspaper.