build log: feb 03 — greenshoot enrichment & analytics

2026-02-03

this build log is automatically generated

projects
dream-backend apprentice-web reducibl seedling-web greenshoot-web gatewaystack.com inner-web
session stats
1.3k tool calls 108 files 64 sessions
tool breakdown
Read
249
Glob
20
TaskCreate
7
TaskUpdate
14
Write
68
Edit
318
Bash
536
Grep
34
Task
5
TaskOutput
13
WebFetch
21
WebSearch
5

what i shipped today

i got the greenshoot enrichment script running smoothly with batch processing and fuzzy matching. the landing page is live on firebase, and i’m tracking waitlist sign-ups. plus, i’m getting a better handle on the competitive landscape.

batch processing saves the day

the enrichment script was crashing with “out of memory” errors. turns out, loading 1,033 plant profiles and all their associated pdfs and docxs at once is a lot. i rewrote the script to process plants in batches of 50, and memory usage is now stable. a simple fix, but a crucial one.

fuzzy matching for dundee data

the dundee edible plants data wasn’t merging with the usda data because of scientific name mismatches. i implemented fuzzy matching using the string-similarity library, and now i’m getting a 100% match rate on the first batch. it’s amazing how much difference a little fuzziness can make.

landing page is live

the greenshoot landing page is now live on firebase. i connected it to the existing joinwaitlist cloud function, and i’m tracking sign-ups. i also added plausible analytics to the page, so i can see how people are using it.


interested in working together? let's talk

← back to build logs