Project 1 · Preview
Snapshot from 01 Jun 2026 — sent by the tracker to my inbox, untouched.
✈️ Daily Flight Price Report — 01 Jun 2026
Trip to New Haven
| Route | Date | Flight | Stops | Duration | Price | 30d median | 30d low |
|---|---|---|---|---|---|---|---|
| Detroit → New Haven | 20 August 2026 | 401 (Avelo) | Non-stop | 1h 39m | $78 below median |
$86 | $78 |
Round Trip: Kolkata ↔ Mumbai
| Route | Date | Flight | Stops | Duration | Price | 30d median | 30d low |
|---|---|---|---|---|---|---|---|
| Kolkata → Mumbai | 16 September 2026 | 1136 (Akasa Air) | Non-stop | 2h 45m | ₹5,610 below median |
₹6,566 | ₹5,610 |
| Mumbai → Kolkata | 20 September 2026 | 1107 (Akasa Air) | Non-stop | 2h 55m | ₹5,610 below median |
₹5,614 | ₹5,610 |
It watches several trips at once, in dollars and rupees, and only writes when something lands below the 30-day median.
Project 2 · Live preview
The real app, running here with sample balances. Try the recommender — change the amount or category.
Home
Tell me about an upcoming purchase and I’ll pick the card that earns you the most.
Business class to Paris
You’re 75% there.
Sample balances, real logic: the live app syncs from AwardWallet and asks Claude to sanity-check every redemption rate.
Project 3 · Two frames
From a real script the engine generated out of a management research paper — one protagonist, Sarah, carried across every shot.
Shot 01 · Cold open
“Meet Sarah. She manages a local retail branch, and honestly? She’s drowning. Her inbox is a war zone…”
Shot 09 · The payoff
“Investing in managers yields a staggering 847% ROI over three years — turning them into force multipliers, not bottlenecks.”
Frames rendered by the engine from its own storyboard; narration verbatim from the generated script.
Get in touch
Tell me what you are working on. I read everything.
This opens your mail app with the message ready, addressed to saanchi.agarwal@yale.edu. Prefer to write it yourself? That address works too.