Basic risk tools — no account required
An AI-powered business risk and intelligence platform: a fraud transaction scanner, a business financial health score, and the research methodology that connects them. Try our basic checks free, no sign-in required. Create a free account anytime for deeper insights — CSV uploads, portfolio scoring, model explainability and saved history.
Fraud Detection
Check a single transaction for signs of fraud — no account needed.
Business Failure Prediction
Five figures, one score — a quick read on a company's financial health.
What each tab includes
Fraud Detection
Check a single transaction for signs of fraud. No account required. Sign in for the full transaction scanner — CSV upload, an entire file scored at once, and model explainability behind every result.
Basic Fraud Risk Check
Check a single transactionBusiness Failure Prediction
Type in five figures and get a health score out of 100. No account required. Sign in for the full nine-field model, a failure-probability estimate, and the model comparison, radar and driver charts behind the number.
Basic Business Health Check
Five figures, one scoreHow the platform decides
Seven methods, implemented and not just described: Random Forest, gradient-boosted trees, logistic regression, a small transformer encoder, SHAP, LIME, and the evaluation metrics that hold them all to account. Sign in to see the live run log — real hyperparameters, real held-out metrics, and every scoring job this session has executed.
Risk decisions, made with the reasons attached
This platform helps small and mid-size businesses catch fraud earlier and understand their own financial health, without needing a data science team to interpret the result. Every score comes with the reasons behind it.
Our mission
Financial fraud and business failure rarely announce themselves in advance — by the time either is obvious, the damage is usually already done. This platform exists to surface that risk earlier, in plain language, so a business owner or analyst can act on it instead of discovering it after the fact.
Who it's for
Built for small and mid-size businesses, finance teams and analysts who need a fast, honest read on transaction risk or financial health — without paying for enterprise fraud tooling or hiring a data science team to interpret a black-box score.
How it works
Two products, one platform. The fraud scanner derives twelve behavioural features from every transaction — timing, channel, cross-border activity, velocity — and scores each one through a Random Forest and gradient-boosted ensemble. The business health score turns a company's financials into four transparent pillars — profitability, liquidity, solvency, growth — plus a machine-learned failure-probability estimate. Both come with SHAP and LIME explanations, so every score names the reasons behind it rather than asking for trust.
What makes it different
The models are real, not page text — Random Forest and gradient-boosted trees are implemented from scratch and run in the browser every time the platform scores, alongside a logistic baseline kept deliberately visible for comparison. No transaction or financial data ever leaves the browser to reach a server, which keeps the product outside PCI-DSS scope and removes any server-side breach surface for that data entirely.
Development Team
Three co-developers, one platform — from research methodology to the working product.
Istiaque Mahmud
Contributes to research integration, AI/ML model development, system design, testing, and implementation, with a focus on fraud detection and business failure prediction.
imr.rafat@gmail.com
Dipon Das Rahul
Contributes to AI/ML implementation, platform development, system integration, testing, and data-driven fraud and business risk assessment features.
dipondasrahul@gmail.com
Sayer Bin Shafi
Contributes to platform development, AI/ML implementation, system integration, testing, and deployment of fraud detection and business failure prediction features.
sayershafi@gmail.comAbout the Team
The three co-developers are MBA graduates of Midwestern State University (MSU Texas) who collaborated on the design, development, testing, and implementation of the platform. The team focuses on transforming research-based methodologies into practical AI-driven tools for financial fraud detection and small-business failure prediction.
Get in touch
Questions about the platform, the research behind it, or setting up a full account? Reach any of the co-developers directly.
Already have an account? Sign in and reach the team from your dashboard for account-specific questions.
Sign In