← Back
Live demo · Deal predictor

See how the model I built for Cotality could score your pipeline.

Adjust the deal parameters on the left. Win probability, risk tier, and the top drivers update live. Everything runs in your browser — no data leaves this page.

Deal inputs

Deal size $50,000
Sales stage Discovery · Proposal · Negotiation
Stakeholders on deal 3
Products in deal 2
Days in pipeline 30 days
Rep talk-to-listen ratio 45%

Prediction

Win probability
42%
±8pp confidence band · Model v1 (demo)
Medium

Top drivers

Recommendation

Balanced deal — no strong outlier. Keep engagement cadence steady.
How this works

Not a toy. A miniature of a shipped tool.

The scoring logic in this demo is a lightweight logistic model with coefficients directly inspired by the four findings I isolated during my Cotality capstone: multi-product attach (H1), talk-ratio effect on win rate (H2), engagement cadence (H3), and Solution Engineer involvement (H4).

The real Cotality build added ML explainability (SHAP), a bootstrap-ensemble confidence interval, and a live upload path so leaders could re-score their pipeline the moment fresh CRM data landed.

What a hiring team gets, in ~3 weeks:

  • A predictive model trained on your deal history
  • An executive dashboard with per-deal drivers
  • A live "what if" simulator like this one, built into the tool
  • An in-app AI chatbot answering stakeholder questions from the data
  • Documentation, tests, and a hand-off any analyst can maintain

I do the analytics and the interface. That’s the whole point.

Next step

Bring this to your team.

I'm interviewing for full-time Data Science, Analytics, and Product Analytics roles starting August 2026. If your team is thinking about pipeline analytics, sales-engineering coaching, or any “we have the data, we just can't see the pattern” problem — this is what I do.