At Capvance I was on the go-to-market side of investment banking, working on sell-side and buy-side mandates. My job came down to two things: make the pipeline easy to read, and keep outreach performing once the volume went up.


The deal teams were running on spreadsheets and memory. Lists went stale fast, every personalized line was written by hand, and if you asked which campaign had started a conversation, nobody could really tell you.
Mostly internal tooling. I built AI-assisted tools and React dashboards on top of our CRM data, covering pipeline tracking, campaign orchestration and the analytics the team checked every day.
Data came in from government filings, Crunchbase, Apollo and LinkedIn, and went through Clay for waterfall enrichment, dedupe and validation. Nothing reached the CRM until it had been through all of that. Only records above the scoring bar went into a sequence, and the sequence depended on the signal: a company that had just announced a deal got a different opener than one that had gone quiet.
Replies were routed as soon as they came in, and every step wrote back to the CRM. That's the reason the dashboards stayed accurate - they were reading from a CRM that was actually up to date.