Show HN: We built the fastest spreadsheet

Original Article ↗ Hacker News Discussion ↗

A new cloud-based spreadsheet called Row Zero promises Excel- and Google Sheets–style workflows with a Rust/Apache Arrow engine that claims to be 100–1000x faster on large datasets. Commenters probe its technical design (columnar storage, CRDT-based collaboration, Python and pandas integration) and compare it with tools like DuckDB, Polars, Equals, and Excel’s own new Python features. Much of the debate centers on real-world adoption barriers—Excel compatibility, keyboard shortcuts, desktop vs. browser UX, security/IAM and compliance, and the risk of relying on a SaaS service for critical, often sensitive, business data.

Performance & Architecture

  • Marketed as 100–1000x faster than desktop Excel, based on supported row counts, import speeds (e.g., Snowflake/S3), and large lookup operations.
  • Some operations are claimed to be orders of magnitude faster due to better algorithms; a few edge cases reportedly only ~10x faster.
  • Built on a Rust backend with Apache Arrow, columnar storage, and custom CSV parsing; all compute runs in AWS via a websocket command model.
  • Uses S3 + DynamoDB for storage/orchestration and CRDTs for multiplayer editing.
  • A few users report slow or failing imports for specific CSV/JSON cases; team attributes this to parser edge cases and is iterating.

Positioning vs Existing Tools

  • Aimed at users who hit Excel/Sheets limits but want a familiar spreadsheet UI, not a full BI tool.
  • Compared to Excel Online, Google Sheets, Power BI, Equals, DuckDB, ClickHouse, Polars/DataFusion: pitched as “spreadsheet-first,” big-data capable, and complementary to Arrow-based tools.
  • Some see it as a potential Excel distribution partner/acquisition target; others question competition given Excel’s ubiquity.

Excel Compatibility & UX

  • Strives for Excel-compatible formulas and hotkeys; missing functions/behaviors are added on request.
  • Known gaps: some range-update semantics, formula copy-to-range behavior, iterative calculations, array formulas, richer formatting/custom number formats, advanced filtering options, and Excel “Alt + ribbon” shortcuts.
  • Uses row index 0, which some like and others note as a compatibility quirk.
  • Debate over browser vs native desktop UX: many power users insist on offline, native, keyboard-centric behavior and local file workflows.

Security, Compliance & Deployment

  • All workloads run in AWS; supports S3 via IAM role assumption and offers enterprise dedicated hosting.
  • SOC 2 and HIPAA: same technical handling across tiers, but formal reports/BAA only for business plans; clarified after pricing-page confusion.
  • States that customer data is not used for any purpose beyond the service itself.
  • Concerns raised about: granting AWS-account access to buckets, static DB creds, lack of full OAuth/IAM coverage, NSL-proof guarantees, SSO limited to higher tiers, and lack of self-hosting for security-sensitive orgs.

Collaboration, Integrations & Roadmap

  • Link-based read-only sharing without accounts; editing still requires accounts, by design for business reasons.
  • Planned/desired features: write-back to databases (Snowflake/Postgres), push-based updates or webhooks, embedding sheets as dashboards, richer conditional formatting, comments, R support, Python-driven charts (e.g., matplotlib), AI-assisted formulas/code, multi-tenant SSO, and possibly desktop/local-first variants.
  • Early adopters report it “feels really fast and intuitive,” especially for multi-GB financial datasets, but many note adoption hinges on Excel parity, security posture, and desktop availability.