SaaS Products
Sagacito
Situation
A media sales rep pricing perishable inventory — airtime and page space — used manual discounts and disconnected pre-sales tools, leaking margin in seasonal spikes. Houses needed higher yield without blocking deals. Existing workflows could not both recommend a price and stop harmful discounting when data quality was weak.
My mandate
Critical decision
Situation
Decide whether AI pricing auto-commits a rate or recommends a price with explicit override and approval paths when confidence is low.
Options considered
Fully autonomous price write-back
Sales would reject a black box, and weak data could lock harmful rates into live inventory.
Manual pricing with AI as a hidden report
Would leave discount leakage untouched — the original revenue problem.
Recommend with confidence, human override, Pgov guardrails on non-compliant discounts
Observation
Sales behaviour showed margin loss from discretionary discounts during spikes. Model confidence varied with audience and seasonality data quality — the interface had to show recommendation vs decision clearly.
Insight
Human override preserves trust but reintroduces discount risk — so Pgov auto-approves compliant deals and escalates heavy discounts instead of hiding the AI.
Response
Ymax recommends the highest acceptable price from inventory, seasonality, and audience signals; reps can override; Pgov escalates non-compliant or heavily discounted proposals.
Result
Pre-sales-to-revenue pipeline adopted by major Indian media houses (company-context clients). Pricing stayed explainable under weak data instead of silently decisive.
How the system changed
Designed Ymax to surface recommended price against inventory constraints, seasonality, and audience data — with confidence visible enough that weak data did not look certain.
Automated proposal and product-mix bundling so reps could meet campaign needs without defaulting to blanket discounts.
Built Pgov approval-guardrail workflow: auto-approve compliant deals, escalate harmful discounting.
Unified TV seconds and print centimetres into a blended portfolio model, and wired RevX prospecting signals into Ymax.
Outcome
business
Implemented by major Indian media conglomerates including Hindustan Times, Ananda Bazar Patrika, and PVR Cinemas.
organisational
Ymax, Pgov, and RevX function as one connected pre-sales-to-revenue system rather than three tools.
operational
Sales see AI pricing as a recommendation with override and approval paths when data quality is weak.
Scale figures describe the operating context. Personal contributions are stated separately.
Frames
What I would change now
I would show one priced deal end-to-end with weak-data and strong-data variants side by side, so hiring managers see exactly how confidence and overrides behave.
Have a similar challenge?