# Promethium Financial Model AI-powered data analytics platform that eliminates ETL by virtualizing data in-place, enabling business users to query across any data source via natural language. - Canonical: https://finamodel.com/startups/promethium - Excel download: https://finamodel.com/startup-models/promethium.xlsx - Category: Enterprise/Security - Model type: SaaS ARR / Valuation - Funding round: Series A - Funding: $26M - Founded: 2022 - Geography: US (team and patent are US-based; no explicit geography stated) - Customer: B2B ## About the company Promethium is an AI-powered analytics platform that virtualises data in place, allowing business users to query across sources with natural language rather than moving data through conventional ETL pipelines. The company sells annual enterprise subscriptions, with a possible consumption overlay tied to usage. Its value proposition centres on reducing data-engineering friction while making distributed enterprise data more accessible to business users. The model follows new-logo bookings through ARR, billings, revenue recognition, expansion, and churn. ACV remains the primary driver, with a separately stated usage sensitivity; delivery costs, sales capacity, product investment, and cash needs complete the forecast. ## 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 - No-code, no-ETL analytics platform. Users ask questions in natural language; Promethium generates datasets via data virtualization without moving data. - Single UI covering Discover → Prepare → Query → Analyze workflow, replacing a 9-step process (search → access → pipeline → move → transform → dataset → query → visualize → return). - Connects to 200+ data sources (data lakes, warehouses, databases, SaaS apps, business apps) without physically moving data. - AI/NLP/ML layer converts natural language questions into data-driven answers. - Patented technology: US Patent 11,074,252 B2 - NLP-driven search-to-data-answer. - Claims "up to 100x" speed vs. traditional approach. - Each data engineer does "1000X more" productivity per customer case study. ## Market - "85% say data fact-based decision making is a top priority" - "Two-thirds say decision making is only somewhat or rarely data-driven" - "58% of companies base 50% of decisions on gut feel" ## Revenue model Not explicitly stated in deck. Inferred from product and enterprise sales motion: - Enterprise SaaS subscription - annual/multi-year contracts; pricing likely per-seat or platform tier, given enterprise buyer (CDO/data team) and deployment against large data estates. - Potential consumption/usage component given 1.1M queries and 600B rows of data referenced - could support usage-based pricing overlay. - Sales motion: direct enterprise sales (CDO/CTO buyers) with data engineer and business user as end-users. - Channel: direct only implied; no partner/reseller channel mentioned. ## Traction & metrics All from slide 7 (Hypothesis Validation / dashboard), covering ~9 months of customer usage: - 14,714 questions answered with Promethium in 9 months - 1.1 Million queries run with Promethium - 600 Billion rows of data processed - 9.1 PB of data under trusted access Traction chart (slide 11): Bar chart titled "Product Market Fit" showing quarters Q1 21 → Q2 21 → Q3 21 → Q4 21 → Q1'22 with clear upward trend (each bar materially larger than prior). No Y-axis labels or absolute values shown - chart is directional only. GTM status: Referenceable customers, repeatable sales motion, buyer/user type identified. No ARR, revenue, customer count, NRR, or churn figures in deck. ## Competition / moat - No named competitors. Competitive positioning shown via bubble chart (slide 13) with axes "Data Accessible" (Y) and "Ease of Use" (X) - Promethium positioned top-right (highest on both). Competitors unlabeled. - Moat claims: patented NLP-to-data technology (US 11,074,252 B2); no-ETL architecture as structural differentiator vs. traditional BI/ETL stacks (Tableau + Informatica, Snowflake + dbt, etc.); 200+ native connectors. - "Only solution built for fast decision making in today's hybrid & distributed world." ## Team & funding ask / use of funds Team (all from slide 12): - Kaycee Lai - CEO & Founder; GM $120M P&L at EMC; President Waterline Data (acquired by Hitachi Vantara); VP Sales Virsto (acquired by VMware), Delphix, Avamar (acquired by Dell EMC). - Puneet Gupta - VP Product; Director of Product at Workday, Salesforce; PM at Amazon, Informatica; Solutions Architect at MuleSoft. - Ravi Kasamsetty - VP Engineering; VP Engineering Milestone; Sr. Director Eng at Oracle, Responsys. - Brett Arnott - VP Marketing; GM Apptio; VP Product & Marketing Digital Fuel (acquired by Apptio); PM VMware. - Team tagline: "Serial Data Entrepreneurs" with five prior exits collectively. --- ## Recommended financial model - **Archetype + why:** Enterprise SaaS ARR model. Revenue is recurring subscription (annual contracts, enterprise); usage metrics suggest a consumption overlay is possible but primary driver is ACV/ARR. Classic enterprise SaaS build: new logo bookings → ARR → churn/expansion → billings → revenue recognition. - **Forecast horizon & granularity:** 5 years (2022–2026); quarterly for Year 1–2, annual thereafter. Model should be built from Q1 2022 given deck is November 2021 and traction is mid-ramp. - **Key drivers & assumptions:** | Driver | Value | | -- | -- | | Starting ARR (Q1 2022) | Unknown | | New logos per quarter | Unknown | | Average ACV | Unknown | | Gross revenue retention | 85–90% | | Net revenue retention | 110–120% | | Gross margin | 65–75% | | Sales cycle | 3–6 months | | Sales headcount ramp | Seed → Series A scale | | R&D as % of revenue | 30–40% | | S&M as % of revenue | 40–60% | | G&A as % of revenue | 10–15% | | Queries per customer per month | ~122K | - **Scenarios (Base / Bull / Bear - which variables flex):** - **Base:** Steady logo adds, NRR ~115%, gross margin 70%, S&M at 50% of revenue through Year 2 then declining. - **Bull:** Faster enterprise penetration (Fortune 500 CDO mandates); NRR expands to 130%+ as more data sources connected; gross margin improves to 78%+ as virtualization infra becomes more efficient. - **Bear:** Long sales cycles slow logo adds; churn risk if ETL-free value proposition isn't sticky once pilots complete; competitive pressure from Databricks / Snowflake building NLP layers natively. - **Required sheets / outputs:** 1. Assumptions - all drivers in one place, clearly tagged vs. 2. ARR Bridge - beginning ARR + new + expansion – churn = ending ARR, by quarter 3. Bookings & billings schedule 4. P&L (IS) - revenue, COGS, gross profit, S&M, R&D, G&A, EBITDA, net income 5. Headcount plan - by function (Sales, CS, Engineering, G&A) 6. Cash & runway - burn, capital required, implied funding milestones 7. KPI dashboard - ARR, ARR growth %, NRR, gross margin, LTV/CAC (once data available), burn multiple ## Frequently asked questions ### Is the Promethium financial model free? Yes. The Promethium 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.