TigerData (Timescale)
Open-source and source-available core (TimescaleDB under Apache-2.0 plus the source-available Timescale License; pgvectorscale and pg_textsearch open source) monetized via usage-based managed cloud (Tiger Cloud, Ghost) - land developers free, expand to paid production workloads
Only the Feb 2022 Series C valuation (> $1B) was publicly disclosed; Series A and B valuations are undisclosed and the plotted figures are analyst estimates for trail shape only. Rounds themselves are verified (also: seed ~$3.4M in 2017 and a $15M Series A1 led by Icon Ventures in Jan 2019, omitted from the plot for lack of any valuation basis; ~$181M raised lifetime). No primary, secondary, or SPAC transaction found after Feb 2022.
Earnings, margins, COGS & capex
As a private company TigerData discloses only headline metrics. At the Jun 2025 rebrand it claimed 2,000+ customers across 25 countries, over 3 million active databases, mid-8-digit ARR growing >100% YoY, 60%+ gross margins, and 200 people in 25 countries. Total equity raised is ~$181M across 5 rounds, ending with the $110M Series C (Feb 2022, led by Tiger Global, >$1B valuation, alongside Benchmark, NEA, Redpoint, Icon, and Two Sigma Ventures). No Series D, secondary, or S-1 found as of Jul 2026 - a 4.5-year gap since the last priced round. The 2022 Series C release cited 20x revenue growth over the prior two years from 500+ paying customers. No audited financials exist publicly.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~40¢ is cost of goods and ~60¢ operating expense, leaving ~0¢ of operating profit.
Revenue trend
Margins
likely improving with cloud scale and storage/compute separation (fluid storage), but structurally capped by hyperscaler COGS
presumed negative; 200-person headcount against mid-8-digit ARR suggests venture-stage burn
COGS structure
Primarily cloud infrastructure (compute, storage, egress on AWS underpinning Tiger Cloud/Ghost) plus support engineering. The 60%+ gross margin is low-to-mid tier for DBaaS - usage-based database services carry real infra COGS; tiered/object storage and compute-storage separation are the main margin levers.
Capex
Minimal owned capex - infrastructure is rented from hyperscalers (opex/COGS model typical of cloud-native DBaaS vendors).
Latest earnings
n/a
none published
- Customers
- 2,000+ across 25 countries (Jun 2025, company-disclosed)
- Active databases
- 3 million+ (Jun 2025, company-disclosed)
- ARR
- mid-8-digit, >100% YoY growth (Jun 2025, company-disclosed)
- Gross margin
- 60%+ (Jun 2025, company-disclosed)
- Headcount
- 200 people in 25 countries (Jun 2025)
- Total raised / last round
- ~$181M / $110M Series C at >$1B (Feb 2022, Tiger Global)
- Cloud usage growth
- 5x over 18 months to Jun 2025 (paid customers)
Growth drivers
- Agentic AI wave — Ghost (GA Jun 9, 2026) targets AI agents as first-class database consumers - unlimited Postgres databases, fast forking, CLI/MCP-server access, per-query pricing, free tier of 100 compute hours/month + 1TB storage
- PostgreSQL's dominance as the default developer database (most-used DB in developer surveys), which TigerData rides rather than fights
- AI search stack in Postgres — pgvectorscale (vector search) and pg_textsearch (open-source BM25 keyword ranking, v1.0 GA on Tiger Cloud) keep RAG/agent workloads inside one database instead of a bolt-on vector DB
- Expansion beyond time-series — the majority of cloud workloads are no longer time-series (company-disclosed, Jun 2025), widening TAM from IoT/metrics into general operational + analytical Postgres
- Usage-based pricing with product-led land-and-expand from a large open-source community (company-estimated 10-20x the paid base)
Bull & bear
TigerData is one of the best remaining independent pure-plays on 'Postgres eats the database market' plus 'agents become the biggest database users' - a capital-efficient, >100%-growth business with real technical moats that either compounds into a durable platform or gets acquired at a strategic premium in the ongoing Postgres land grab.
- ARR growing >100% YoY at mid-8-digit scale (Jun 2025) - if sustained, ARR plausibly crosses $100M within ~2 years, IPO-track territory
- Ghost + Agentic Postgres shipped GA (Jun 9, 2026) before hyperscalers offered agent-native database primitives; MCP-native access for Claude Code/Cursor/Codex users is a genuine first-mover wedge
- Databricks paid ~$1B for Neon (a younger serverless-Postgres startup at far smaller revenue scale) and Snowflake bought Crunchy Data for ~$250M in 2025 - comparable strategic scarcity value supports the neighborhood of TigerData's 2022 mark in an M&A scenario
- Decade of hard-to-replicate Postgres internals engineering (hypertables, columnar compression, continuous aggregates, fluid storage/forking) - not a thin control-plane wrapper
- Only ~$181M raised - clean cap table by unicorn standards, less preference-stack overhang than over-raised peers
A ZIRP-vintage unicorn mark, no new money in 4.5 years, mid-tier gross margins, and a crowded field where every giant now owns a Postgres asset - TigerData risks being squeezed between hyperscaler defaults and better-funded developer darlings, with the agentic pivot still unproven as a revenue engine.
- The >$1B valuation dates to Feb 2022 peak multiples; comparable private software marks reset sharply post-2022 - a down round or structured raise is a live risk if it must raise
- Mid-8-digit ARR after ~9 years of selling implies a long grind, and 'mid-8-digit' is company-marketing language with no audit - the actual figure could sit at the low end (~$30M)
- 60%+ gross margin caps the multiple the business can earn; usage-based DBaaS margins improve slowly and hyperscaler egress/compute costs are not in its control
- Competitive encirclement: Aurora/AlloyDB/Azure PG by default, Neon-inside-Databricks and Crunchy-inside-Snowflake in the enterprise, Supabase ($10.5B, Jun 2026) in the dev-tool funnel, ClickHouse/InfluxDB in analytics/time-series - every flank is contested
- Rebrand + pivot cadence (time-series leader -> 'modern Postgres' -> 'agentic Postgres' in ~18 months) can read as narrative-searching; agent database spend is still a small, speculative budget line in 2026
- No disclosed path to profitability; if growth decelerates below ~50% the private-market clearing price compresses sharply
What it is worth
Last priced round anchor + ARR-multiple scenarios + strategic-M&A comparables (no public financials; all scenario values are analyst estimates, not marks)
~$400-700M
growth decelerates below 50%, gross margin stalls near 60%, and a needed raise or sale clears materially below the ZIRP-era unicorn print
~$0.8-1.5B
growth persists at 60-100%, ARR reaches the $50-100M range by 2027, valuation roughly holds the 2022 mark on today's tighter multiples
~$2-3B
ARR sustains >100% growth toward $100M+, Ghost captures the agent-database category, or a strategic bidding contest (hyperscaler/Databricks-class acquirer) prices scarcity
Anchor: $110M Series C at >$1B post (Feb 2022, Tiger Global). Company-disclosed mid-8-digit ARR (~$30-70M implied) growing >100% (Jun 2025). High-growth infra-software private comps cleared roughly 10-20x forward ARR in 2025-26 (e.g. ClickHouse at $6.35B, Supabase at $10.5B); strategic Postgres M&A (Databricks-Neon ~$1B for a younger, smaller-revenue asset; Snowflake-Crunchy ~$250M) sets a scarcity reference in an acquisition scenario. Figures are estimates for framing only - not financial advice.
SWOT
Strengths
- Deep Postgres engineering brand — TimescaleDB is one of the most widely deployed Postgres extensions; cofounder CTO Mike Freedman (Princeton CS professor) anchors technical credibility
- Rides PostgreSQL rather than competing with it - compatible with the entire Postgres tool/ORM/driver ecosystem
- Early, differentiated agentic positioning — Agentic Postgres + Ghost with native MCP support for Claude Code, Cursor, Windsurf, Codex, Gemini CLI, and VS Code shipped ahead of most incumbents
- Capital-efficient unicorn — ~$181M raised for a mid-8-digit-ARR business growing >100% - not an over-raised burn machine
- Large open-source funnel (community estimated 10-20x the paid base) feeding product-led growth
Weaknesses
- 60%+ gross margin trails best-in-class DBaaS (MongoDB Atlas-style economics) - usage pricing passes through hyperscaler costs
- No new primary capital since Feb 2022 — the $1B mark is a ZIRP-era print that may not hold today (funding-gap overhang)
- Rebrand risk — 'Timescale/TimescaleDB' brand equity built over a decade partially reset to an unknown name in Jun 2025
- Sub-scale vs hyperscalers and vs MongoDB/Databricks/Snowflake war chests in a consolidating Postgres market
- Revenue still concentrated in a managed-cloud SKU whose open-source core competitors can partially replicate
Opportunities
- Agent-driven database consumption could multiply database-instance counts (each agent spinning ephemeral DBs) - Ghost is purpose-built for exactly this, and Databricks says 80% of databases provisioned on Neon were agent-created
- Postgres consolidation wave (Databricks-Neon ~$1B and Snowflake-Crunchy Data ~$250M, both 2025) makes TigerData one of the few remaining independent Postgres platform assets - strategic acquisition premium
- RAG/search consolidation into Postgres (pgvectorscale + pg_textsearch BM25) displaces standalone vector/search databases for mid-market workloads
- Enterprise migration off Oracle/proprietary DBs to Postgres continues to expand the addressable base
Threats
- Hyperscalers bundling — AWS Aurora/RDS, Azure Database for PostgreSQL, Google AlloyDB own the default procurement path and can clone features
- Databricks (via Neon) and Snowflake (via Crunchy Data) now push serverless/enterprise Postgres with vastly larger GTM engines
- Supabase's developer-mindshare flywheel ($10.5B valuation after its Jun 2026 $500M Series F) competes for the same new-project Postgres developer
- Open-source free-riding: users can self-host TimescaleDB or use community extensions without ever paying
- If the agentic-database thesis matures slower than hoped, the 2025-26 repositioning spend yields little differentiated revenue
Moats, dependencies & bottlenecks
Moats
columnar compression, continuous aggregates, fluid storage/forking) Years of kernel-level work that hyperscaler wrappers and thin serverless layers cannot quickly copy; partially eroded by Postgres core absorbing features over time
Open-source community + installed base of TimescaleDB deployments Large funnel and switching friction for production time-series workloads; weakened by Timescale License (TSL) limits pushing some users to alternatives
Production databases rarely migrate; expansion revenue compounds on retained workloads
Real first-mover advantage in Jun 2026 but easily contested - Neon/Databricks and Supabase are racing at the same agent-database use case
Dependencies
infrastructure supplier Tiger Cloud runs on hyperscaler infrastructure (AWS-first); AWS is simultaneously its largest cost line and a direct competitor via Aurora/RDS/Timestream
technology foundation The entire product is a Postgres superset; roadmap depends on upstream community direction it does not control
Cursor, OpenAI Codex, etc.) distribution channel Ghost's wedge assumes MCP-style agent tooling remains the standard interface; protocol shifts would strand integration work
No primary raise since Feb 2022; continued >100% growth investment likely requires new capital or M&A within the planning horizon
Advantages
- Genuine technical differentiation inside the world's most popular open-source database rather than beside it
- First credible 'database for agents' story shipped to GA with hard spending caps and MCP-native UX
- Capital efficiency and a >100% growth rate that most 2015-vintage infra startups lost long ago
- Scarcity value as one of the last independent Postgres platform companies after the 2025 M&A wave
Weaknesses
- 60%+ gross margin trails best-in-class DBaaS (MongoDB Atlas-style economics) - usage pricing passes through hyperscaler costs
- No new primary capital since Feb 2022 — the $1B mark is a ZIRP-era print that may not hold today (funding-gap overhang)
- Rebrand risk — 'Timescale/TimescaleDB' brand equity built over a decade partially reset to an unknown name in Jun 2025
- Sub-scale vs hyperscalers and vs MongoDB/Databricks/Snowflake war chests in a consolidating Postgres market
- Revenue still concentrated in a managed-cloud SKU whose open-source core competitors can partially replicate
Bottlenecks
- Gross-margin ceiling from hyperscaler COGS limits reinvestment capacity relative to software-margin peers
- Enterprise go-to-market scale: ~200 employees against competitors fielding thousands of sellers
- Brand rebuild post-rebrand: 'TigerData' must re-earn the search/SEO/mindshare equity 'Timescale' held
- Monetizing the free tier — Ghost's generous free plan (100 compute hrs/month, 1TB) must convert without subsidizing agent-spam workloads
Top signals & trends
Top signals
Concrete product shipping against the agentic thesis, ahead of most incumbents
Rare transparency for a private DB vendor; metrics healthy, though self-reported and unaudited
Either efficient growth on existing cash (bullish) or inability/unwillingness to raise at or above the $1B mark (bearish) - unresolved
Validates Postgres platform scarcity value; positions TigerData as a prime remaining target - but also arms rivals
Evidence the TAM expansion beyond the original niche is real, not just narrative
Could be fast market-reading or narrative churn; watch whether Ghost produces disclosed traction by 2027
Trends
strong tailwind · TigerData monetizes Postgres demand directly; every Postgres convert is pipeline
potentially transformative tailwind · If agents multiply database-instance demand (80% of Neon's provisioned DBs were agent-created, per Databricks), agent-native platforms capture disproportionate share; still early
pgvectorscale + pg_textsearch (BM25) position Postgres against standalone vector DBs (Pinecone et al.) and Elasticsearch-class search engines
Raises exit value but strengthens competitors' bundled offerings
Default-path procurement favors AWS/Azure/GCP native Postgres; independent DBaaS must win on product
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Primary cloud infrastructure under Tiger Cloud/Ghost
Upstream open-source database TigerData extends
crypto-fintech, energy, and SaaS observability teams 2,000+ paying customers across 25 countries (Jun 2025); usage-based, mid-market-weighted; most logos not publicly itemized
Publicly documented case study (crypto market data; migrated from InfluxDB to TimescaleDB, 94% storage reduction per the case study)
Default managed-Postgres path for AWS customers; also TigerData's infrastructure landlord
Bundled enterprise Postgres with Azure credits and agreements
AlloyDB targets the same 'faster Postgres for analytics' claim
Neon (acquired ~$1B, May 2025) brings serverless Postgres with fast branching - the closest analog to Ghost's forking model - now backed by Databricks GTM
Snowflake Postgres (via the ~$250M Jun 2025 Crunchy Data acquisition) targets enterprise Postgres inside the Snowflake account footprint
Postgres-based developer platform with dominant new-project mindshare; $10.5B valuation after its $500M Series F (Jun 2026)
Atlas competes for the same operational-workload wallet; different data model but same buyer
Valued at $6.35B in its May 2025 Series C (Khosla-led), extended later in 2025; wins high-scale real-time analytics workloads that outgrow Postgres-based approaches
Legacy head-to-head time-series rival (Messari publicly migrated InfluxDB -> TimescaleDB); less relevant to the new agentic positioning
Postgres-compatible distributed SQL for global OLTP; overlaps at the high end
Compete for AI search workloads that pgvectorscale aims to keep inside Postgres