A production-deployed revenue operations intelligence dashboard built to demonstrate Senior Sales Operations readiness at Sorenson Communications.
This repo is a fork of @munas-git's AI-Powered Sales KPI Dashboard. The original served as the wireframe. This is the build.
The following were stripped out by design:
- MongoDB and pyodbc — replaced with a lightweight CSV-native data layer. Three flat files (pipeline.csv, revenue.csv, reps.csv) remove all infrastructure dependency and make the app fully portable on Streamlit Community Cloud.
- LangChain, RAG agent, and OpenAI chatbot — removed entirely. A live API dependency introduces failure risk in a demo context; the analytical depth of the dashboard carries the demonstration without it.
- Generic sales data — replaced with synthetic data modeled on Sorenson Communications' five service lines, SLA structure, and three-motion revenue operation (New, Expand, Renew) — the analytical patterns of any enterprise interpreting services business, built with Sorenson's specific context by name.
Forking, stripping, and rebuilding a scaffold is a deliberate product judgment call. Every removal was a decision about what actually serves the purpose.
| Tab | Title | Core Business Question |
|---|---|---|
| 1 | Executive Summary | How is the business performing right now at a glance? |
| 2 | Pipeline & Forecast | Which deals are at risk and what will we close this quarter? |
| 3 | Multi-Motion Revenue | How are our service lines and sales motions trending? |
| 4 | Rep Productivity & Compensation | Who is performing, who needs coaching, and are reps paid fairly? |
- Risk-adjusted forecasting — composite deal risk scoring (stage overage, risk flags, deal size) feeding three forecast scenarios: conservative, base, and upside. Built to replace gut-feel pipeline calls with a reproducible, stage-weighted model.
- Pipeline hygiene tracking — at-risk deal table with four boolean hygiene fields (past stage benchmark, high discount, low engagement, manual risk flag) modeled after the fields a Dynamics 365 CRM admin would track.
- Consumption velocity cohort matrix — utilization heatmap showing how much of their contracted interpreting hours each customer cohort consumes by month of tenure. Red cells at month 3 signal churn risk before month 12 renewal.
- Compensation plan simulator — four-tier commission logic (floor, ramp, at-quota, accelerator) with real-time payout calculation at any attainment level, including the floor cliff edge case that generates the most rep disputes.
Three synthetic CSV files in the /data folder. Fixed random seed (42) ensures reproducibility across every run.
| File | Rows | Contents |
|---|---|---|
| pipeline.csv | 135 deals | deal_id, rep, territory, motion, stage, ARR, days_in_stage, risk flags |
| revenue.csv | 480 rows | 8 service lines × 60 months, MRR, contracted hours, utilized hours, SLA met |
| reps.csv | 12 reps | quota, attainment, win rate, avg deal size, days to close, bonus, TAM |
The data is simulated. The logic — formulas, risk scoring, compensation tiers, cohort structure — is production-ready.
| Layer | Technology |
|---|---|
| App framework | Python 3.12 + Streamlit |
| Visualization | Plotly Express + Graph Objects |
| Data processing | pandas + NumPy |
| Deployment | Streamlit Community Cloud |
| Version control | GitHub (auto-deploys from main) |
| Data | CSV (3 files, no database) |
git clone https://github.com/datadynamo-hub/Senior-Sales-Operations-Analyst.git
cd Senior-Sales-Operations-Analystpip install -r requirements.txtstreamlit run app.pyBuilt by Jonathan Khan for Sorenson Communications interview preparation.