Designed and built a scalable data ingestion pipeline in Python, Django, and Procrastinate (PostgreSQL-backed task queue) that automates the reliable processing of complex professional election records (PVs) from raw documents to structured, queryable data. Shipped new features on the benefits.artikle.io platform, including a human-only live support chat with two-way Slack threading, so support staff can reply from Slack and users see replies in-app. Optimized data retrieval from the pipelines to surface company benefits and PV data to end users, including information that was previously not exposed, using scheduled cron jobs and LLM-assisted extraction.

Profile Image

Profile Image




Profile Image

Profile Image




Profile Image

Profile Image




Profile Image



Profile Image

Profile Image


Profile Image

Profile Image




Profile Image

Profile Image
No images available

Profile Image

Profile Image



























Profile Image



Profile Image



Profile Image




Profile Image







Profile Image







Profile Image
No images available
No images available
No images available
No images available

Profile Image




Profile Image






Profile Image



Profile Image









Profile Image







Profile Image


Profile Image













Profile Image






Profile Image

Profile Image



Profile Image
No images available










Profile Image

Profile Image


Profile Image

Profile Image

Profile Image

Profile Image

Profile Image

Profile Image

Profile Image

Profile Image

Profile Image

Profile Image
No images available

Profile Image

Profile Image

Profile Image

Profile Image
No images available

Profile Image

Profile Image

Profile Image
No images available







Profile Image

Profile Image
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available
No images available