shipped field record
Enterprise System Integration
I worked at TalkingPoints from 2022–2026. Each project lists its own dates, so some begin later within that span.
- Problem
- Customer context was fragmented across systems with conflicting identifiers.
- My role
- Designed reconciliation logic and production integrations.
- Shipped
- Four connected systems with fuzzy matching and reviewable customer views.
- Proof · 2026-07
- Unified customer context across Salesforce, PlanHat, Customer.io, and support.
Public systems
Salesforce, PlanHat, Customer.io, Intercom/Dixa. Fuzzy matching with Jaro-Winkler. Philosophy: minimize friction for humans, don't replace them.
Materials & technology
Links
Frequently Asked Questions
Why fuzzy matching instead of exact matching?
Real-world data is messy. School districts use abbreviations, alternate names, and typos. Exact matching would miss most connections. Fuzzy matching with the right threshold finds the matches that matter while avoiding false positives.
What's the hardest part of enterprise integration work?
Data quality and politics, not technology. Each system has its own truth about customers, and those truths don't always agree. The technical work is reconciliation; the harder work is getting teams to trust unified data over their familiar silos.