PARKHI.ai-
Predictive Anomaly Recognition & Knowledge-based Hazard Intelligence

CSV
Raw Upload Ready
EDA
Quality Insights
ML
Risk Scoring

REAL-TIME SIGNAL VIEW · USER · DEVICE · IP · PAYMENT · LOCATION

SCROLL
THE PROBLEM

Fraud hides inside
messy, inconsistent transaction data.

Real transaction datasets arrive with broken amounts, duplicate columns, mixed timestamp formats, missing values, typo-heavy city names, and invalid IP addresses. If the data is not cleaned first, the fraud signals stay buried.

See the workflow →
HIGH RISK TRANSACTION

TXN-104882

Flagged after cleaning and feature generation

94 / 100

Amount deviation12.4x above user average
Device behaviorFirst time device detected
Location mismatchBilling city differs from transaction city
Velocity spike | New device | Location anomaly

Fraud Detection, End To End

From messy transaction files to interpretable fraud predictions, every stage is designed for practical analyst workflows.

Data Cleaning Engine

Standardize raw CSV uploads by fixing amount formats, normalizing timestamps, merging duplicate columns, and correcting invalid records.

EDA Quality Checks

Surface missing values, duplicate rows, transaction_id conflicts, invalid IPs, and distribution shifts before modeling begins.

Behavior Features

Derive average spend, deviation from normal behavior, transaction velocity, unusual transaction time, and location drift for each user.

Device And Location Signals

Track new devices, device reuse, city standardization, and geo mismatches that often reveal account takeover or synthetic activity.

Fraud Risk Scoring

Use engineered features to power a fraud model that returns a prediction and a transaction-level risk score for every record.

Explainable Alerts

Show why a transaction was flagged with feature contribution hints so analysts can trust the decision and act faster.

HOW IT WORKS

From raw transactions to
interpretable fraud insights.

Consulting feeLoan repaymentService invoiceInvestment returnANexus HoldingsBMeridian TrustCAlbatross LLCDPacific Ventures
CIRCULAR FLOW DETECTED
01

Upload

Accept raw transaction CSV files

Bring in datasets containing amount, timestamp, location, device, IP address, payment method, and user activity fields without changing your backend flow.

02

Clean

Standardize messy real-world records

Merge duplicate columns like amt into transaction_amount, normalize timestamps, standardize city names, remove duplicate records, handle missing values, and flag invalid IP addresses.

03

Explore

Run EDA on quality and behavior

Highlight nulls, duplicates, invalid entries, and key distributions across transaction amounts, payment methods, device usage, and time-of-day activity.

04

Model

Engineer features for fraud detection

Create user-level behavioral signals such as spending deviation, transaction velocity, unusual devices, location drift, and odd transaction times before scoring.

05

Explain

Predict risk and show why

Return a fraud prediction, risk score, and explanation layer so users can inspect the model decision instead of treating it as a black box.

DEMO PREVIEW

Preview the fraud workspace
before an analyst investigates.

PARKHI.ai
3 high-risk transactions queued
Live

Flagged Transactions

TXN-104882

User U-219 | Visa

94

TXN-204771

User U-031 | Wallet

83

TXN-778420

User U-558 | UPI

77

TXN-116903

User U-074 | Card

58

TXN-440018

User U-902 | NetBanking

36
CSV cleanedSHAP-ready

Relationship View

USRTXNDEVIPLOCPMTTIMEAVG

Risk score raised by amount deviation, unseen device, and late-night velocity burst

Transaction Detail

TXN-104882

March 23, 2026 | 02:14 AM

User: U-219 | Payment: Visa

City: Bengaluru | Device: Android 14

IP status: Corrected and validated

94 /100
Amount deviationNew deviceUnusual hour

Explore the fraud dashboard →

0

Cleaning and analysis stages

0+

Behavioral fraud indicators

0

Outputs per transaction

0

Explainable risk score

Built for noisy data,
clear fraud decisions, and fast review.

PARKHI.ai | Predictive Anomaly Recognition & Knowledge-based Hazard Intelligence | Tech Sagar 2026

PARKHI.ai

© 2026 | Predictive anomaly recognition workspace

GitHub →