TigerEye

Sharper Vision. Smarter Sales.

Inspiration

Sales reps at SRS Distribution often lose deals without fully knowing why or whether the deal could have been recovered. The hackathon theme “Agents at Work” inspired us to create an AI teammate that helps uncover lost-deal reasons and provides guidance on possible recovery.

What it does

TigerEye is an AI sales recovery agent. It:

Explains why a deal was lost.
Surfaces evidence from activity logs (emails, calls, proposals).
Suggests recovery options and next steps.
Offers inline actions like “Find Alternates,” “Draft Recovery Email,” and “Create Follow-up Task.”
Lives inside the Lead Detail page so reps get context and AI insights in one place.

How we built it

Frontend: Next.js 15, TailwindCSS, shadcn/ui.
Pages/Routes: /dashboard, /chat, /leads/[leadId].
Lead Detail Page: Built demo data for Ashley Taylor @ GMS Inc. with profile, company info, deal summary, and lost reasons.
Chat Integration: Connected directly to the provided API, sending messages and rendering structured responses (reply, reasons, evidence, next_steps).
UI Orchestration: Added a Chat Panel on the lead page with suggestion chips and inline actions.

Challenges we ran into

Parsing structured API responses and displaying them cleanly.
Preserving bullet and newline formatting from the API.
Designing a polished UI in limited time.
Balancing demo data with realistic functionality.
Handling errors and empty states gracefully.

Accomplishments that we're proud of

Delivered a working chat integration with the provided API.
Built a realistic Lead Detail demo showing how AI can recover deals.
Designed a clean Next.js + shadcn UI with chat, tabs, and inline actions.
Created a strong foundation for future CRM and sales tool integrations.

What we learned

The importance of structured AI outputs to make insights actionable.
How to quickly design and prototype UIs with Next.js and shadcn/ui.
That UX orchestration (chat + inline actions) is just as critical as AI logic.
The potential for AI to help reps recover deals that would otherwise be lost.

What's next for TigerEye

Connect to live CRM and order data for real evidence grounding.
Add tools for alternate SKU lookup, pricing guidance, and freight optimization.
Introduce recovery scoring with logistic regression:
$$ P(\text{recoverable}) = \frac{1}{1 + e^{-(\beta_0 + \beta_1x_1 + \dots + \beta_nx_n)}} $$
Expand beyond recovery into deal win predictors, proactive alerts, and coaching for sales reps.