By: Brian Murray, Sean Whitney, Zao Chen
We’re thrilled to be leading ParadeDB’s Series A and supporting them as they build the next generation of search infrastructure on Postgres.
Developers love Postgres. The rise of Postgres as the default relational database has been one of the most significant shifts in developer infrastructure over the past decade. It’s no longer just a solid open-source option; it’s becoming the dominant database for new applications. We’ve seen this firsthand through our investment in Supabase, where Postgres is at the heart of the product. As developers increasingly consolidate their stack around Postgres, it raises a natural question: what other key workloads should now be reimagined in a Postgres-first world?
One of the most important of those workloads is search.
Search remains a mission-critical function for modern apps. From user-facing product search to internal analytics and monitoring, developers need fast, powerful, and customizable search capabilities. Historically, the answer was to bolt on a dedicated search engine like Elasticsearch, OpenSearch, or Algolia. That made sense in a world of heterogeneous databases and microservices. But it also introduced a host of new challenges: data syncing, infrastructure sprawl, and operational overhead.
ParadeDB is built for a new reality — a world where Postgres is the center of gravity.
ParadeDB is a Postgres-native search engine that supports full-text, vector, and hybrid search out of the box. It’s built as a Postgres extension (like pgvector) and designed to feel like a seamless part of your database, not a bolted-on subsystem. It’s open source, developer-friendly, and deeply aligned with how modern teams want to build.
ParadeDB isn’t just stitching existing solutions together — they’ve made deep technical improvements to how search works inside Postgres. Their open-source engine, pg_search, brings powerful search functionality (like ranking, faceting, filtering, and fuzzy matching) into the core of the database. It’s fast, accurate, and keeps data indexed in real time without any of the headaches of managing an external search system. Because of this ParadeDB has been quickly adopted, powering search at scale and in production for companies like Bilt Rewards, Modern Treasury, UnifyGTM, and Alibaba.
The product feedback from these customers is exceptional. Teams consistently point to dramatically faster indexing, improved relevance, and a smoother developer experience compared to Elasticsearch. Just as importantly, they love the operational simplicity of having search live natively inside Postgres, with no ETL, no lag, and no duplicated infrastructure.
We were initially introduced to ParadeDB by a CTO at one of our portfolio companies. When we looked under the hood, we found that the team was already known in the Postgres community. The Supabase team — who have a deep technical bench and high standards — corroborated the ParadeDB team’s chops and were enthusiastic about the product direction. That sealed it.
ParadeDB is led by Philippe Noël and Ming Ying. From our very first conversation, Phil and Ming impressed us with their energy, velocity, and positivity. That combination of technical rigor and founder grit is exactly what we search for (pardon the pun).
Postgres is on a path to dominance. Search is too important to live outside your database. The future belongs to systems that are fast, flexible, and integrated — that’s exactly what ParadeDB is building and we’re excited to be along for the ride.
Why We Invested in ParadeDB was originally published in Craft Ventures on Medium, where people are continuing the conversation by highlighting and responding to this story.