FSConsultory — Business Intelligence Consulting

// 01 — About

About me.

Francisco Santelices

ICI · UDP · Santiago

I'm Francisco Santelices, an Industrial Civil Engineer from Universidad Diego Portales, with over 7 years building Business Intelligence solutions in retail and commercial environments.

My work goes beyond dashboards. What drives me is building tools people actually use — reports that sit on a manager's screen every morning, models that surface fraud before it costs money, and systems that make data feel less like a spreadsheet and more like a decision.

I've worked across retail chains with 600+ locations, financial regulatory environments, and multinational operations — always focused on turning complexity into clarity. Today I integrate AI agents into analytical workflows to push that clarity even further.

Years of experience

600+

Operational locations

50%

Faster reporting

Expert & advanced tools

Masters · Power BI · Python · AI

Countries · Multinational projects

// 02 — Work

Nothing invented. Nothing assumed.

CLIENT PROJECT

Terraza Abril · Alhué, Región Metropolitana

Restobar · Parrilla & Sushi

Smart Assistant for Restaurant Management.

A WhatsApp assistant that logs costs, answers questions about the business — sales, costs, tips — and leaves a trace of everything it claims.

n8n WhatsApp Business API AI Agents Python Google Sheets Fudo API

● Delivered · Sep 2026

First FSConsultory client

THE PROBLEM

A restaurant owner knows the day's sales. The margin is only known at month's end — when nothing can be done with that information anymore.

Costs do not arrive on their own: they live in crumpled receipts, in loose messages and in purchases nobody got around to logging. No system solves that while logging them remains administrative work competing with the service.

WHAT IT DOES

Cost logging

The team logs a purchase by speaking, typing or photographing the receipt. The assistant shows what it understood, the person confirms or corrects in plain language, and only then is it saved. Everything lands in the management software and in a backup sheet, with the original transcript available for auditing.

Business analysis

Answers sales, expenses by category, best-selling products and margin, with charts readable on a phone propped on the kitchen counter. Permissions are set by phone number: who can log and who can query.

Tip system IN DEVELOPMENT

Next phase: distribution and traceability of tips per shift, under the same verification criteria as the rest.

HOW IT WAS BUILT

🔍

Every reading is verified against the document.When processing a receipt, the system compares the sum of the line items against the printed total and flags a mismatch instead of saving it.

❓

Ask rather than assume.Faced with a purchase that does not look like the business, the system shows what it understood and asks. Silently deciding which expense counts is worse than interrupting.

✅

Never say done when nothing was saved.No confirmation is issued before the operation is actually written. A false positive on an expense record is an invisible loss.

📊

A correct number can still be a wrong answer.If the weekly meat purchase falls on a single day, the daily chart suggests a loss where there is none. The message was adjusted so true data cannot lead to false conclusions.

Anyone can build a WhatsApp bot. The difference lies in the decisions you never see: what the system does when it is not sure, and what it refuses to claim.

— FSConsultory · Engineering criterion

Now available for other restaurants, with Fudo and other point-of-sale systems.

See the assistant → Book a 30-min demo

POWER BI PROJECTS

Real projects developed for companies across different industries and countries.

01 / RETAIL

01 / RETAIL BI

Multi-Branch Retail Dashboard

Solución Power BI end-to-end con ETL en Power Query, modelamiento en estrella y capa semántica DAX. Mide throughput operacional (pedidos/hora), tasa de cumplimiento de entregas, evolución de ganancia vs período anterior (AA) y rentabilidad por SKU de categoría. Publicado en Power BI Service con RLS por sucursal.

Power BI DAX Power Query ETL RLS Star Schema YTD / MTD

$6.7M

BILLING

BRANCHES

3.000

ORDERS/YEAR

02 / COMMERCIAL BI

Commercial Analytics Platform

Sistema de BI con modelo relacional de alta cardinalidad (800+ clientes, 12+ estados). Implementa medidas de inteligencia de tiempo con DATEADD y CALCULATE para comparativos período actual vs LY. Incluye visualizaciones geoespaciales con Bing Maps, tooltips enriquecidos y drill-through entre módulos. Segmentación de clientes por status y clasificación Class-A/Class-B con lógica DAX condicional.

Power BI DATEADD CALCULATE Drill-through Bing Maps RFM High Cardinality

MODULES

800+

CLIENTES

12+

ESTADOS

02 / COMMERCIAL BI

03 / MULTINATIONAL XSales Dashboard

03 / MULTINATIONAL BI

Multinational Sales Analysis — XSales

Snowflake schema adapted for 5-country operations. Country → Customer Type → Product drill-down hierarchy. Calculated measures for marginal contribution per SKU, ROI by market and discount elasticity. Responsive Mobile/Desktop layouts via bookmarks.

Power BI Snowflake Schema Drill-down P&L Analysis Bookmarks Margin Analysis Responsive

$118.7M

BILLING

COUNTRIES

13.2%

MARGEN

04 / SALES TRACKING

Sales Tracking Dashboard — Colombia

Star schema operational dashboard (sales facts + store, salesperson, product, date dimensions). DAX measures for market share by branch, category rotation speed and salesperson benchmark vs team average. Conditional treemap and geographic map of Colombia.

Power BI RANKX TOPN Star Schema Bing Maps Treemap Benchmark DAX

BRANCHES

MED·BOG·BUC

CITIES

04 / SALES TRACKING Colombia Sales Dashboard

View Full Portfolio →

// 03 — Experience

Professional Journey.

08/2019 — Present

Senior BI & Loss Prevention AnalystCURRENT

Empresas SB

Architected end-to-end BI solutions in Power BI (star schema, DAX, RLS) consumed daily by C-level and operational teams — cutting reporting cycles from days to hours.
Built anomaly detection models in Python + SQL monitoring loss and fraud signals across 600+ nationwide retail locations, strengthening internal control significantly.
Designed the national commercial KPI model (RANKX, TOPN, CALCULATE) aligning salesperson incentives with real business outcomes — deployed across all branches.
Optimized inventory management through stockout and overstock analytics, reducing dead stock and improving category replenishment cycles.
Led the national bonus calculation process, ensuring full traceability and precision across all commercial areas.
Monitor retail-supplier commercial agreement performance through KPI analysis (Loss/Purchase ratio), identifying deviations and providing data-driven evidence to support contract renegotiations.

Power BISQLPythonExcel

2018
2 months

Professional Internship

SBIF — Modeling Dept.

Work carried out in the Modeling and Methodology Department. Research, creation and study of probabilistic models to understand phenomena in the Chilean banking sector, laying the foundations of Basel II (the fundamental banking regulatory framework of the Basel Committee on Banking Supervision).

Probabilistic ModelsBanking

2017
2 months

Operations Internship

ICIL-ICAFAL Maquinaria S.p.A

Supported machinery rental, maintenance and fleet management processes.

Operations

// 04 — Services

How can I help you?

01 / 📊

Executive Dashboards

Power BI reports connected to real data. Leadership sees the business in real time, no spreadsheets.

02 / 🔍

Operational Control & Loss Prevention

Identify deviation, fraud and inefficiency patterns. Proven across 600+ locations.

03 / ⚙️

Custom Data Models

SQL, Python and Excel models automating repetitive analysis and generating actionable indicators.

04 / 🤖

AI Agents

Develop and integrate AI agents into analytical and operational processes, boosting decision quality.

05 / 💬

Data Communication

Translate complex analysis into clear language for executive, commercial and operational teams.

06 / 🎯

Commercial Optimization

Incentive models and KPIs aligned with real business outcomes through data and AI.

HOW I WORK

🔎

Diagnosis

Understand your business, data and the real problem.

✏️

Design

Define architecture, key metrics and deliverables.

⚙️

Development

Build models, dashboards or agents.

🚀

Implementation

Deploy in your real environment.

🎓

Training

Transfer knowledge for team autonomy.

Turning data
into executive decisions.

About me.