Case study · Automation
Automation that opened eight new markets
A business-analyst & automation internship at Goodera (remote, Feb–Jun 2022): Python scrapers over web and email sources, feeding cleaned, deduplicated imports into the CRM — so market research stopped being a copy-paste job.
Fig. 01 — Scattered sources funnelled into clean CRM rows: the whole internship in one picture.
02 — The problem
Growth research that burned analyst hours
Expanding into new markets meant researching partners and opportunities by hand: find the sources, copy the details, paste them into the CRM, repeat. The information existed — scattered across websites and inboxes in inconsistent formats — but collecting it consumed hours that should have gone into judgement, not transcription.
Sources were scattered and inconsistent — no two websites or mail threads shaped alike.
The CRM needed clean, deduplicated imports — garbage in would poison outreach.
A remote, distributed team: the workflows had to run without me in the room.
03 — Process
Collect, clean, hand over
STEP 01 · Collect
Python scrapers for web and email
Purpose-built collectors pulled partner and market data from web pages and email sources — 47,500+ data points extracted and classified over the internship instead of retyped.
STEP 02 · Clean
Validation and dedupe into the CRM
Import pipelines normalised, validated and deduplicated the raw haul before it touched the CRM, keeping the outreach database trustworthy as it grew.
STEP 03 · Hand over
Workflows the team could run alone
The collectors and imports were documented and packaged so the team kept running them after the internship ended — the ~1,200 saved hours kept compounding.
Fig. 02 — Pipeline detail figure coming soon (internal tooling; the numbers are from my internship record).
04 — Outcome
Hours back, markets open
The automation mapped 8 new markets and 14 partner organisations while saving roughly 1,200 hours of manual collection — and it set the pattern I've reused ever since: scrape carefully, validate hard, and give the team a workflow instead of a spreadsheet.
Next project