Pecunio AI — Financial Model Platform
Confidential — Pecunio AI Platform Overview
Pecunio • AI Financial Model Platform

Platform Demonstration

Enterprise Financial Modelling with AI-Powered Analysis

March 2026  |  Mock data — Global Consumer Health Division

01 — Platform Overview

How Pecunio Works

Connect your accounting system. The platform builds a financial model. An AI agent answers questions over it.

Data Sync
ERP / GL → PostgreSQL
Accounts, journals, bank txns
→
Model Seed
Raw data → nodes, series,
edges, category mapping
→
AI Agent
33 tools, LangGraph,
reasoning + narrative
→
Calc Engine
Deterministic ops,
cell provenance
→
Validation
11 reconciliation checks,
tieouts, variance
Key principle: The AI agent handles reasoning, narrative, and tool selection. A deterministic calculation engine handles all financial math — no LLM touches any number. Every cell traces back to source data through an auditable provenance chain.
02 — Financial Model

CEO Summary — Model View

The model auto-generates from synced accounting data. Users interact via AI agent or direct tab navigation.

Global Consumer Health Division — Financial Model

12 months actual (Jan – Dec 2025)  |  Connected to ERP  |  Base Scenario

Guide
CEO Summary
P&L
Cash Flow
Balance Sheet
Revenue
P&L Explorer
AR / AP
KPIs
Variance
Scenarios
Health
Knowledge
Ops
$847.2M
Revenue
+8.3% YoY
$142.6M
Net Income
16.8% net margin
62.4%
Gross Margin
+1.2pp vs prior
$224.8M
Forecast Run-rate
Next 3 months
38 days
DSO
-2 days vs prior Q
+3.2%
Revenue MoM
Dec vs Nov 2025
83.2%
Cost Ratio
-0.8pp improvement
$528.8M
Gross Profit
12 months actual
$34.7M
Open AR
$8.2M overdue >60d
4 ops
Active Operations
2 drivers, 2 overrides
8 PASS 2 WARN 1 FAIL Revenue ties to GL · Cash ties to bank · Monthly P&L reconciled
Performing Well
  • Revenue +8.3% YoY driven by OTC Pharma (+14.2%)
  • Gross margin expanded 120bps on procurement savings
  • Operating leverage improving — OpEx grew only 6.1%
Needs Attention
  • Medical Devices COGS spiked 22% in November
  • $142K OpEx variance vs GL — likely 3 unsynced journals
  • Consumer Wellness declined 3.1% (seasonal softness)
03 — Virtual CFO Agent

AI Agent Over the Financial Model

The Virtual CFO opens as a floating panel over the live model. Users ask questions while the financial data stays visible underneath.

My Financial Model Base 6-Month Avg
+ What-If ↻ Sync from Xero ↧ Export Excel
54 pass 1 warn 16 fail
Guide
CEO Summary
P&L Explorer
Xero P&L
Cash Flow
Balance Sheet
AR
Revenue
KPIs
Variance
Scenarios
Health Check
Knowledge
P&L Explorer ○ Draft
P&L Section Q1 FY24
Aug–Sep		 Q2 FY24
Oct–Dec		 Q3 FY24
Jan–Mar		 Q4 FY24
Apr–Jun		 Q1 FY25
Jul–Sep		 Q2 FY25 F
Oct–Dec		 Annual
Total
▾ Revenue001,336,1751,892,4412,841,1082,817,3839,770,107
▾ Contract Fees received0(6,474)1,336,175740,2291,089,71403,781,122
▾ Unmapped0(6,474)1,336,175740,2291,089,714—3,781,122
▾ Client A Hospitality001,438,281762,8081,089,714—3,891,336
Client A Hospitality001,438,281762,8081,089,714—3,891,336
▹ Release Fees06,474010,8422,776—65,199
▹ Client Refunds0000(801)—(2,470)
▹ Interest Income000428134—434
▹ Sale of Uniform Income000642520—1,559
▹ Other Revenue00000—0
▹ Client Fees00001,723—35,170
▹ Cost of Sales0420993,9311,404,2082,108,5142,063,0175,889,052
Gross Profit04202,330,106488,233732,594754,3667,522,045
Gross Margin %0.0%0.0%174.4%179.6%175.3%175.3%177.0%
▹ Other Income0995,2305,792112264,0631,595,614
▹ Operating Expenses(125,079)416474,591430,29221,376300,2712,603,224
Net Profit(125,079)9352,809,9264,876,59189,2552,306,56021,490,990
Net Profit %0.0%0.0%210.3%197.3%232.1%232.1%220.0%
EBITDA(125,079)8362,804,6964,870,80089,1432,042,49719,895,376
Virtual CFO ▾
⤢ × ↕
how can we improve margins starting in April?
MW

• Big lever: margins improve fastest by either lifting price or reducing delivery cost per $ of revenue (your COGS is largely payroll-like, so small efficiency gains matter).

• Starting April, I’d focus on 4 moves (in order):

  1. Make the April price rise “stick” without volume loss: implement a clean uplift on Client Fees only (exclude interest/other income), and pair it with a value narrative + 30-day notice. If even 1–2% volume churn shows up, it can erase a chunk of the margin win.
  2. Tighten gross margin through workforce utilisation (best for payroll-heavy COGS): target higher billable utilisation, reduce bench time, and reprice low-margin roles. Practically: minimum bill rate floors, tighter overtime controls.
  3. Review operating expense run-rate: your OpEx jumped in recent quarters. Identify which costs are one-off (system implementation, onboarding) vs recurring. Target OpEx/revenue ratio below 30%.
  4. Customer mix shift: grow higher-margin segments. Your top client is 41% of revenue — diversifying into mid-tier accounts typically yields +2–3pp margin improvement.
model a 5% price increase on Client Fees from April
MW
vnext_preview_ops

Here’s the impact of a 5% growth driver on Client Fees for Apr–Jun:

Before (Q4 baseline)$762,808
After (+5% driver)$800,948
Delta+$38,140/quarter

Net margin impact: +0.4pp. Should I commit?

📎
↑
×
How this works: The agent has 33 registered tools covering model views (P&L, cash flow, balance sheet), forecasting (preview/commit ops, baselines, scenarios), analysis (revenue dashboard, KPI metrics, variance), drilldown (ledger lines, customer concentration), quality checks (11 reconciliation tests), and knowledge integration (document upload, assumption extraction). The agent panel floats over any tab — the user can ask questions while viewing live model data.
04 — Financial Model

P&L Statement — Actual + Forecast

4 ops applied · Baseline: avg_6m · Click any cell for provenance

Global Consumer Health Division — Financial Model

6 months (Oct 2025 – Mar 2026)  |  3 actual + 3 forecast

Guide
CEO Summary
P&L
Cash Flow
Balance Sheet
Revenue
P&L Explorer
AR / AP
KPIs
Variance
Scenarios
Health
Line Item Oct 25 Nov 25 Dec 25 Jan 26 F Feb 26 F Mar 26 F Total
REVENUE
OTC Pharmaceutical Sales$52.8M$54.1M$56.7M$58.4M$60.2M$62.0M$344.3M
Medical Device Revenue$22.6M$21.9M$22.3M$22.3M$22.3M$22.3M$133.7M
Consumer Wellness$10.8M$11.2M$12.5M$11.5M$11.5M$11.5M$69.0M
Licensing & Royalties$2.2M$2.2M$2.5M$2.3M$2.3M$2.3M$13.8M
Total Revenue$88.5M$89.4M$94.0M$94.5M$96.2M$98.1M$560.7M
COST OF SALES
Raw Materials & API($14.8M)($15.1M)($15.7M)($15.2M)($15.2M)($15.2M)($91.2M)
Manufacturing & Packaging($8.5M)($8.7M)($9.1M)($8.8M)($8.8M)($8.8M)($52.7M)
Quality & Compliance($3.4M)($3.5M)($3.6M)($3.5M)($3.5M)($3.5M)($21.2M)
Distribution & Logistics($5.2M)($5.4M)($5.6M)($5.4M)($5.4M)($5.4M)($32.5M)
Total Cost of Sales($32.0M)($32.7M)($34.0M)($32.9M)($32.9M)($32.9M)($197.6M)
Gross Profit$56.4M$56.7M$60.0M$61.6M$63.3M$65.1M$363.1M
OPERATING EXPENSES
Sales & Marketing($12.5M)($12.8M)($13.2M)($12.8M)($12.8M)($12.8M)($77.1M)
Research & Development($8.6M)($8.6M)($9.1M)($9.5M)($9.5M)($9.5M)($54.9M)
General & Administrative($4.8M)($4.9M)($5.0M)($4.9M)($4.9M)($4.9M)($29.6M)
Regulatory & Legal($2.8M)($3.0M)($3.1M)($3.0M)($3.0M)($3.0M)($17.8M)
Total Operating Expenses($28.8M)($29.4M)($30.5M)($30.2M)($30.2M)($30.2M)($179.4M)
Net Income$27.7M$27.3M$29.5M$31.3M$33.1M$34.9M$183.7M
Forecast (baseline) Override / driver applied Negative value
Cell Provenance: Click any forecast cell to see the full calculation chain — source data, baseline method, operations applied, knowledge document reference, and approver. Every value is traceable to its origin.
05 — P&L Explorer

Customer Concentration — Revenue by Segment

Same P&L structure, pivoted by customer instead of account. Interactive treemap and drilldown table.

14.7%
Top-1 (CVS Health)
Single-client share
52.9%
Top-5
Concentration
64.5%
Top-10
Concentration
824
HHI Score
Low concentration
#CustomerSegmentRevenueMargin% TotalMoM
1CVS HealthEnterprise$124.8M63.0%14.7%+4.2%
2Walgreens Boots AllianceEnterprise$98.2M62.0%11.6%+2.8%
3McKesson CorpEnterprise$78.4M61.0%9.3%+1.9%
4AmerisourceBergenEnterprise$62.1M61.0%7.3%−1.2%
5Cardinal HealthEnterprise$54.6M61.0%6.4%+3.5%
6Kaiser PermanenteHealth Systems$42.8M63.0%5.1%+6.1%
7Rite AidRetail$28.4M60.0%3.4%−8.2%
8HCA HealthcareHealth Systems$24.2M62.0%2.9%+1.4%
9Walmart HealthRetail$18.6M60.0%2.2%+12.4%
10Cigna GroupInsurance$14.8M58.0%1.7%+2.1%
All Other (142 customers)$300.4M61.5%35.5%
Total$847.2M62.4%100%
Portfolio is well-diversified: HHI of 824 indicates low concentration. No single customer exceeds 15% of revenue. The top-5 contribute 52.9%, a healthy distribution across major enterprise accounts.
Watch list: Rite Aid revenue declining −8.2% MoM — potential credit risk. AmerisourceBergen also trending down (−1.2%). Customer concentration treemaps available via the P&L Explorer tab.
06 — Cash Flow & Receivables

Cash Flow Statement & AR Aging

Direct method cash flow with 3-month forecast. AR aging analysis with collection risk flags.

$130.2M
Current Cash
Dec 31, 2025
$14.6M
Avg Net Cash
per month
38 days
DSO
-2 days vs prior Q
$8.2M
AR Overdue >60d
3 customers flagged

AR Aging Buckets

CustomerTotalCurrent1–30d31–60d61–90d90+d
CVS Health$8.2M$5.1M$2.1M$620K$360K—
Walgreens$6.4M$4.0M$1.4M$580K$390K—
McKesson$4.8M$2.8M$1.3M$420K$280K—
AmerisourceBergen$3.6M$1.8M$940K$480K$240K$160K
Rite Aid$2.8M$840K$620K$512K$480K$360K
Total AR$34.7M$18.4M$8.1M$4.2M$2.8M$1.2M
Collection risk: Rite Aid has $360K in the 90+ day bucket — highest single-customer overdue balance. AmerisourceBergen also aging at $160K 90+ days. Both flagged for follow-up. Remaining portfolio is well within terms.
07 — Platform Capabilities

Feature Summary

Complete feature set of the Pecunio AI financial modelling platform.

Deterministic Forecasting Engine

  • Op-based: every change is a discrete, reversible operation
  • 5 driver types: growth, fixed, ratio, allocation, step change
  • 6 baseline methods: trailing averages, last month, YoY
  • Effective date ranges on all operations
  • Category-level drivers (apply to all accounts in a group)
  • Cell-level provenance: traces every value to source data + logic

Preview / Commit Approval Workflow

  • Agent recommends changes with before/after diffs
  • Human reviews impact before anything persists
  • Every committed op tracked: who created it, when, from what source
  • Undo any operation without side effects
  • Agent-created vs human-created ops distinguished
  • Analogous to SOX “read → recommend → approve” pattern

Data Quality & Reconciliation

  • 11 automated checks: revenue ties, expense ties, cash balance ties
  • Section-level + account-level variance analysis
  • Known variance suppression (acknowledged discrepancies)
  • KPI snapshot integrity audits
  • Uploaded report cross-validation vs live data
  • NaN/infinity guards throughout the calculation pipeline

Knowledge Base & AI Context

  • Upload PDF, Word, Excel as business context
  • LLM extracts structured forecast assumptions from documents
  • Each assumption: title, rationale, confidence score, proposed op
  • Propose → approve → reject → commit workflow
  • Per-tab AI executive commentary generation
  • RAG context layers for policy/SOP-aware agent responses

33+ Agent Tools

  • Model views: P&L, cash flow, balance sheet, summary KPIs
  • Forecasting: preview ops, commit ops, change baseline, scenarios
  • Analysis: revenue dashboard, KPI dashboard, variance report
  • Drilldown: ledger lines, top contacts, customer concentration
  • Quality: reconciliation, accuracy audit, report validation
  • Context: knowledge docs, assumption extraction, executive briefing

Visualization & Reporting

  • Schema-driven rendering: backend defines visual structure
  • 7 block types: charts, KPI grids, tables, status lists, key-value
  • Dual chart engine: Recharts + ECharts (treemaps, composed charts)
  • Conversation-aware rendering: auto-show/suppress based on intent
  • Customer concentration treemaps with segment grouping
  • 27-component React workspace with 13+ tabs

Technology Stack

Backend

Python · Django · DRF
LangChain · LangGraph
PostgreSQL · 27 tables

Frontend

React · TypeScript · Vite
shadcn/ui · Tailwind CSS
Recharts · ECharts · SWR

Infrastructure

GCP Cloud Run · Cloud Build
OAuth2 API Integration
CI/CD · Auto-migrations

Learn how we helped 100 top brands gain success