About Dreambase 💭

Dreambase is the AI-native analytics platform for Supabase. We connect directly to your Supabase database and let AI agents generate dashboards, reports, and insights automatically — no data team, no pipelines, no SQL required. We are a small, fast-moving team building something genuinely new at the intersection of AI, analytics, and the Supabase ecosystem.

Your Mission

Build the analytics engine that makes Dreambase fast, accurate, and scalable across any data source and any volume. You will own the query layer, the OLAP processing pipeline, and the on-the-fly transformation systems that turn raw data from Supabase, ClickHouse, PostHog, and external APIs into production-ready dashboards, reports, and insights — at startup speed and enterprise scale.

What you'll do

• Own the Query and OLAP Layer — Design and implement the systems that execute analytical queries across heterogeneous data sources. You know the difference between OLTP and OLAP workloads and you architect for both.

• Build On-the-Fly Transformations — Implement DuckDB and PGlite-powered transformation pipelines that execute in Edge Functions and WASM environments. Calculations, aggregations, and dataset joins that happen at the edge without hitting production databases.

• Scale Across Data Sources — Build the connectors and query routing layer that handles Supabase Postgres, ClickHouse, PostHog, and external APIs as unified analytics sources. Different engines, one consistent output.

• Design the Caching Architecture — Build the intelligent dataset caching system that persists query results at the right layer — Supabase Storage, edge cache, or in-memory — based on data freshness requirements and query cost.

• Power the Reports Engine — Implement the data pipeline that generates executive summaries and scheduled reports enriched from multiple sources, with transformations and calculations handled correctly at every layer.

• Optimize Relentlessly — Profile slow queries, design partition strategies, and build the monitoring systems that catch performance regressions before they reach users. You think about query plans the way other engineers think about code reviews.

• Own the Analytics Tools — Build the Query Analyzer, Dataset Analyzer, and the transparency layer that shows users exactly what is running against their data sources and how it performed.

What we're looking for

• OLAP Expertise — You have built analytical query systems at scale. You understand columnar storage, vectorized execution, query planning, and the architectural differences between OLAP and transactional workloads.

• Distributed Systems Depth — You understand MapReduce patterns, distributed aggregation, and the tradeoffs involved in processing large datasets across partitioned data. You have worked with these systems in production, not just read about them.

• DuckDB and In-Process Analytics — You have used DuckDB in production and understand why in-process OLAP changes the calculus for analytical workloads. Bonus if you have explored PGlite for client-side or edge query execution.

• ClickHouse Experience — You have built or maintained ClickHouse-backed analytics systems and understand its data model, performance characteristics, and operational tradeoffs at scale.

• Edge and WASM Deployment — You understand what it means to deploy analytical compute to Vercel Edge Functions, Cloudflare Workers, or WASM environments. You know the constraints and you design systems that respect them.

• PostHog Familiarity — You understand PostHog's data model and event schema well enough to build reliable connectors and cross-source analytics on top of it.

• Postgres Depth — Comfortable with query planning, indexing strategies, and performance optimization in Postgres, including Supabase-specific patterns like PostgREST and RLS-aware query design.

• Accuracy Standards — A wrong number in an analytics product is a critical bug. You build validation systems and measure output correctness before calling anything done.

Bonus points for

• Production experience with ClickHouse, Apache Druid, Pinot, or other purpose-built OLAP databases

• Experience with Apache Parquet, Apache Iceberg, or columnar storage formats at scale

• Familiarity with streaming analytics pipelines and real-time aggregation systems

• Experience running analytical workloads in WASM or edge compute environments

• Background as a data engineer or analytics engineer at a high-growth SaaS or developer tools company

• Open-source contributions to DuckDB, PGlite, or analytics tooling

• Experience building multi-source analytical systems that unify structured and event-based data

Why this role is different

Most analytics engineering roles are about moving data from one warehouse to another. This one is about building the compute layer that runs analysis as close to the data as possible — in edge functions, in WASM, in DuckDB — so that Dreambase can deliver sub-second insights across enterprise-scale datasets without requiring a data warehouse, a dedicated data team, or a six-month implementation project.

If you have wanted to work on the hard part of analytics infrastructure, this is it.