Synlitics — Restaurant Analytics SaaS

01 — Problem

Product Context

Restaurant operators managed data across disconnected tools — POS systems, delivery platforms, review sites, and spreadsheets. Weekly reporting consumed 5+ hours of manual work. Beyond the time cost, operators had no early warning system for business risks like declining repeat rates or delivery margin erosion.

02 — System Architecture

Data & System Design

Built a complete analytics platform: automated data ingestion from multiple sources, centralized warehouse, automated reporting, and proactive alert system.

POS System
Sales data

→

Delivery APIs
Swiggy / Zomato

→

n8n ETL
Automated ingestion

→

Supabase / PostgreSQL
Central warehouse

→

Power BI
Dashboards + Alerts

03 — Platform

Key System Components

Data IngestionAutomated ETL via n8n workflows: POS data, delivery APIs, Google reviews — all into Supabase/PostgreSQL

Data ModelingStar-schema warehouse: fact tables for transactions, dimension tables for products, time, and segments

ReportingPower BI dashboards auto-refreshed daily: revenue trends, item performance, delivery comparison, retention cohorts

Alert SystemAutomated alerts for anomalies: revenue drops >15%, review decline, delivery cancellation spikes

Churn DetectionRule-based churn risk scoring for repeat customers using recency/frequency analysis

04 — Output Layer

Dashboard & Reporting

Power BI dashboards replaced manual reporting entirely. Operators see key metrics at a glance with daily auto-refresh and trend analysis.

Weekly Revenue

₹4.2L

Avg Rating

4.3★

Repeat Rate

32%

At-Risk

127

05 — Insights

Key Findings

Delivery platform margins were 15-20% lower than dine-in — operators were unaware of per-channel cost
Repeat customer rate was the strongest predictor of long-term revenue stability
Negative reviews correlated with vendor changes — supply chain signal in review data
Weekend sales were highly predictable — enabling better scheduling and inventory positioning

06 — Decision Layer

How the platform drove decisions

Operators moved from "monthly look-back" to "daily awareness." The alert system surfaced problems before they compounded. One operator used channel margin analysis to renegotiate delivery commissions, saving ₹50K/month. Another adjusted weekend staffing 30% based on demand patterns.

07 — Impact

Business Impact

Reporting Time

-5 hrs/wk

Fully automated

Churn Detection

Proactive

Early warning system

Data Sources

5+

Unified warehouse

08 — Learnings

Reflections

"The hardest part wasn't the technology — it was understanding the operator's workflow. We initially built beautiful dashboards, but operators wanted alerts and summaries. Redesigning around 'tell me what's wrong' instead of 'explore your data' dramatically increased adoption."