# Rasgo Financial Model ML feature store platform that lets data scientists transform, share, and serve features to production models - built natively on top of cloud data warehouses (Snowflake, BigQuery, Redshift). - Canonical: https://finamodel.com/startups/rasgo - Excel download: https://finamodel.com/startup-models/rasgo.xlsx - Category: AI/ML - Model type: SaaS ARR / Valuation - Funding round: Series A - Funding: $20M - Founded: 2021 - Geography: US-based (no explicit geo stated in deck). - Customer: B2B ## About the company Rasgo provides a feature store built natively on cloud data warehouses including Snowflake, BigQuery, and Redshift. Data teams can transform, share, version, serve, and monitor production features without maintaining a separate warehouse. The platform claims a 140-times reduction in feature-compute cost, 17-times faster queries, and a 30-minute Snowflake deployment. It had five paying clients, 70,000 downloads of its open-source pyrasgo library, 13 employees, and $25.1 million raised after 11 months. The model converts open-source users and enterprise sales opportunities into subscription ARR, then tracks contract expansion, churn, warehouse usage, delivery costs, and gross margin. It layers sales capacity, engineering and customer-success hiring, operating cash flow, cash burn, and runway. ## What's included - 5-year monthly revenue build with stage-appropriate growth assumptions - Full P&L, headcount plan, and operating-expense schedule - Cash-flow statement, runway, and burn-rate tracking - Valuation via exit multiple with a DCF cross-check - Returns analysis with MOIC and IRR - Unit economics including CAC, LTV, payback, and cohort retention ## Product & value proposition - Feature store that sits on top of existing cloud data warehouses (ELT architecture, not a separate warehouse). - Three core workflow pillars: - **TRANSFORM**: Profile data, calculate feature importance, apply transforms via click or code using the Rasgo transform library. - **SHARE**: Version/track features, enable discovery and reuse across data science teams, expose lineage to business stakeholders. - **SERVE**: Promote features to production in a single click/command; monitor data drift; integrate natively with MLOps and data pipelines. - Open-source Python library (`pyrasgo`) as a community/top-of-funnel entry point. - Key technical differentiators: 140X reduction in cost to compute; 17X+ faster feature query performance; 30-minute Snowflake deployment; dev features immediately production-ready. ## Revenue model Not explicitly stated in deck. Inferred from context: - Enterprise SaaS subscription, likely seat-based or consumption-based, given the enterprise customer profile (Fortune 500 companies, energy firms, financial trading firms). Rationale: all named customers are large enterprises; the product integrates into existing cloud data warehouses suggesting per-seat or usage-tier pricing is standard in this space. - Open-source `pyrasgo` library (70k downloads) likely serves as a PLG (product-led growth) funnel into paid enterprise contracts. Rationale: common pattern for ML infrastructure tools (dbt, Great Expectations, etc.). - Distribution channel: direct enterprise sales (CEO background in GTM at Platfora, Kinetica, Domino Data Lab, Confluent). ## Traction & metrics All from slide 2 - "11 Month Highlights": - **Investment**: $25.1M total raised (Insight Partners & Unusual Ventures) - **Team growth**: 330% - grew from 3 to 13 employees - **Paying clients**: 5 paying clients secured - **Open-source downloads**: 70k pyrasgo downloads - **Product validation**: 400+ user interviews; won head-to-head battles vs. competition Named enterprise clients/validators: - AES (energy - Head of Data Science + Analytics quoted) - Farmer's Fridge - Guardant (health) - Stanley Black & Decker - An unnamed F500 manufacturing company - An unnamed financial trading firm (Christian Peressin, Managing Partner) - Snowflake Technology Alliance partnership (Director of Technology Alliances quoted) No revenue figures (ARR, MRR, ACV) disclosed. ## Competition / moat Competitive landscape: - Positioned as disrupting legacy ETL/ELT tooling: Alteryx, Informatica, Trifacta, Paxata (data engineering side) and fragmented data science tooling (Apache Spark, pandas, scikit-learn). - Integrated with (not competing against) cloud data warehouses: Snowflake, BigQuery, Redshift; and workflow tools: dbt, Apache Airflow. Moat claims: - ELT-native architecture eliminates duplicate data warehouse - cited 140X cost reduction, 17X query speed. - Feature versioning, experiment time travel, lineage tracking - stickiness through institutional ML knowledge lock-in. - OSS community (pyrasgo, 70k downloads) as distribution moat. - Snowflake Technology Alliance validation and co-sell potential. ## Team & funding ask / use of funds **Founders**: - **Jared Parker** - CEO; GTM leadership background at Platfora, Kinetica, Domino Data Lab, Confluent. Owns marketing, sales, alliances, finance, ops. - **Patrick Dougherty** - CTO; MS in Analytics; data scientist by trade; built data science practice at Slalom from ground up. Owns product and engineering. **Funding**: $25.1M total raised. No current ask amount, round size, or use-of-funds breakdown disclosed in deck. --- ## Recommended financial model - **Archetype + why**: Enterprise SaaS ARR model with a PLG (product-led growth) conversion layer. The business has a dual-motion: open-source `pyrasgo` downloads drive organic awareness → direct enterprise sales convert to paid contracts. Revenue is almost certainly ACV/ARR-based subscription given the enterprise customer base. SaaS ARR model is the right archetype. - **Forecast horizon & granularity**: 3 years (Year 1–3), monthly for Year 1, quarterly for Years 2–3. Early-stage company with only 5 paying clients - monthly granularity matters for cash planning in Year 1. - **Key drivers & assumptions**: - **New logos per quarter**: 5 paying clients in ~11 months as baseline; initial ramp of 1–2 new enterprise logos/quarter, accelerating to 3–5 by Year 2. Rationale: small direct sales team, enterprise sales cycles typically 3–6 months. - **Average Contract Value (ACV)**: $60k–$120k/year per enterprise client. Rationale: early-stage B2B ML infrastructure; F500 names suggest ability to command 6-figure deals, but 5-client base implies pricing is still being established; comparable tools (Tecton, Feast enterprise) typically price in this range. - **Net Revenue Retention (NRR)**: 110–120%. Rationale: feature store becomes deeply embedded in ML pipelines; expansion natural as more data science teams onboard; no churn data in deck but enterprise ML tooling typically has high retention once embedded. - **Gross logo churn**: 5–10% annually. Rationale: early-stage product risk; no deck evidence of churn direction. - **Gross margin**: 70–80%. Rationale: SaaS software with cloud infrastructure costs; ELT-native architecture (no separate compute to run) should yield high margins vs. managed compute alternatives. - **Headcount / S&M**: 13 employees at time of deck; S&M grows in line with new logo targets; assume 2–3 quota-carrying AEs in Year 1 at ~$150k OTE each. - **R&D spend**: ~40–50% of OpEx in Year 1 given early product stage and CTO-led team. - **PLG conversion rate (OSS → paid)**: 0.1–0.3% of OSS downloads convert to a paid trial/pipeline. Rationale: 70k downloads → 5 paying clients is roughly 0.007% at deck date, but many downloads are likely individual hobbyists; enterprise conversion funnel is separate from raw download count. - **Cash burn / runway**: derive from OpEx build; $25.1M raised suggests significant runway, but headcount growth (330% in 11 months) implies aggressive burn. - **Scenarios (Base / Bull / Bear - which variables flex)**: - **Bull**: New logo adds 4–5/quarter by Year 2; ACV scales to $150k+; NRR reaches 125%; Snowflake co-sell accelerates pipeline. - **Base**: New logo adds 2–3/quarter by Year 2; ACV $80k; NRR 115%. - **Bear**: Sales cycles elongate; only 1–2 new logos/quarter; ACV stays at $50–60k; churn 10%. - Flex variables: new logo adds, ACV, NRR, gross margin. - **Required sheets / outputs**: 1. **Assumptions** - all drivers with scenario toggles. 2. **ARR Bridge** - beginning ARR, new ARR, expansion ARR, churned ARR, ending ARR. 3. **P&L (Income Statement)** - Revenue, COGS, Gross Profit, S&M, R&D, G&A, EBITDA, Net Income. 4. **Headcount Plan** - by department (S&M, R&D, G&A), with loaded cost. 5. **Cash Flow & Runway** - cash burn, months of runway given current raise. 6. **KPI Summary** - ARR, logo count, ACV, NRR, gross margin, burn multiple, CAC payback. 7. **Dashboard** - charts: ARR bridge waterfall, ARR vs. headcount, logo growth, gross margin %. ## Frequently asked questions ### Is the Rasgo financial model free? Yes. The Rasgo model is a free Excel download with live formulas. ### Can I change the assumptions? Yes. The workbook is editable and its live formulas recalculate when assumptions change.