Perplexity agent spent $2.20 screening 327 land parcels and produced a 4K site-report video
garrytan · x · 2026-11-20
A user asked Perplexity to find land for a 12-cabin build near Sacramento. The agent called Mireye's Land Search & Enrich API — 31 requests for $2.20 — narrowing 327 parcels to 3 finalists by filtering out >15% slopes, flood zones, wetlands and protected land.
It then pulled USGS 3DEP 1m elevation and NAIP aerial imagery, modeled each site in three.js, recommended a winner, and produced a 4K 60fps video recapping the process. Both Perplexity's CEO and Garry Tan amplified it as a striking example of agent orchestration across external APIs.
More from coding & agent
- How a 50-customer B2B app automated QA with agents simulated from production MLflow traces — Lopsided_Feedback_81 · 2026-11-20
- SpaceX SI's talk on shipping 11,000 production PRs/month hits 1M views fastest — film_girl · 2026-11-20
- Glean on serving enterprises at the model Pareto frontier: it comes down to evals — verrsane · 2026-11-20
- LATO raises $4.7M for AI agents that do primary research, publishes data center report — nathanbenaich · 2026-11-20
- Dev slams Google Antigravity as weak vs Claude Code and Codex — jasondeanlee · 2026-11-20
- 'Sandbox is the new harness': agent isolation becomes the hot frontier — pjausovec · 2026-11-20