Taralytix
Build v7.2.3 · UI reliability hotfixLocal model · summarized AI context

Upload source files

Select several Excel or CSV files. Each workbook sheet becomes a source that can contribute data to the shared model.

Drop files here

Revenue, cost, volume, forecast, or operating data can be separate files.

Map source columns

Taralytix auto-maps common finance headers. Confirm the dimensions and measures before building the model.

Upload files first.

Validate and build the evidence model

Confirm measure coverage, joins, duplicates, and granularity before AI analyzes the results.

Run validation after mapping your files.

Customize report

Set the company identity, reporting period, logo, and visual style before generating the analysis.

Report identity

Company logo

LOGO

PNG, JPG, WebP, or SVG. Maximum 2 MB. The logo appears once in the report header, not inside every chart.

Professional theme

Select analyses

Start with Taralytix Smart Recommendation or apply a professional report template. Every template is filtered to analyses your mapped data can support.

0 selected0 available
Smart Recommendation

Let Taralytix choose the most decision-useful analyses

Build the model to receive a data-aware recommendation.

Manual controls
Choose at least one analysisYour selections and chart styles are saved in this browser tab.
Taralytix V7.2

AI Finance Command Center

Checking AI…

One grounded AI workspace replaces manual narrative pages. It reads the current deterministic model and produces an executive brief, driver analysis, risk scan, actions, and management-ready reporting.

Numbers stay deterministic.Taralytix calculates; AI explains, challenges, and writes.
No user setup.The app calls its own secure /api/analyze endpoint.
Evidence first.Missing values, assumptions, and uncertainty must be disclosed.
Live model scope

Evidence sent to AI

Build a data model to activate AI.

Checking the secure AI backend…

One-click analysis

Ask for a focused review

Ask anything

Question the model

Ctrl/⌘ + Enter to send

Your financial intelligence brief appears here

Build the model, choose the current scope, and generate a complete report. Users never enter an API key or Worker address.

Direct answerExecutive summaryKey findingsDriversRisksActions

Selected analysis report

Only the analyses chosen in the report builder are shown and included in Print / Save PDF.

Executive overview

Portfolio results from the shared filtered model.

Trend

Cost composition

Regional scorecard

Operating efficiency

Location-level cost and revenue productivity.

Cost per unit by location

Portfolio benchmark

Location detail

Period change

Compare two years or periods using the same dimensions and measures.

Regional change

Business mix

Revenue and volume concentration by business unit and category.

Revenue mix

Mix summary

Business detail

Exceptions

Rule-based signals derived from the current filtered model.

Priority issues

Scenario simulator

Stress revenue, volume, and cost drivers using a volume-linked cost model without changing uploaded data.

Inbound, outbound, and warehouse costs scale with volume. This setting controls how much remaining cost also scales with volume.

Baseline vs scenario

Unified data model

Aggregated canonical records used by every analysis page.

Canonical record preview