Nanonets Financial Model

• Platform ingests documents (PDFs, images, email attachments, paper) via API, email, cloud storage connectors (Google Drive, Dropbox, S3, OneDrive, SharePoint).
• AI/ML extracts structured data (OCR + deep learning fine-tuned per customer).
• Learnable decision engines execute post-extraction workflows: PO matching, ERP writes, approvals, payments, CRM updates, Zendesk ticketing - all configurable via natural language prompt.
• Key differentiator: self-learning ML that improves accuracy with more data (vs. rule-based OCR that degrades); customer model fine-tuned from base model via human feedback loop.
• Primary use cases: AP automation (invoices → ERP), customer support (claims/tickets), procurement, HR onboarding.
• Vision: "5-person $1B company" via full workflow orchestration across applications.

• Global industrial automation software market: $33.54B (2020) → $60.83B (2028), ~2x growth.
• Intelligent Document Processing (IDP) sub-market: CAGR of 36.8% during forecast period.
• RPA market breakdown cited: Finance & Accounting 25%, Supply Chain 15%, HR 15%, IT 15%, Customer Support 10%, Sales & Marketing 5%, Other 15%.
• Within Finance RPA: AP automation 33%, Payroll 27%, AR 13%.
• 34% of Fortune 500 have already used the product.

• Core unit: per-page / per-extraction pricing.
• Entry price: $0.1/page.
• Freemium → paid conversion; scales with volume.
• Land-and-expand motion:
• Land: $0xx/extraction (entry per-page rate, exact cents redacted in deck).
• Expand - More Volume: $0.xx/extraction (volume discount tier).
• Expand - More Data Types: $xk/year (annual subscription layer for additional document types).
• Expand - More Automation: $0.0x/step/workflow (per-step fee for workflow automation).
• Net effect: ACV grows as customer automates more workflow steps and document types - typical upsell path from extraction-only to full workflow orchestration.
• Channels: PLG (freemium self-serve); content marketing (SEO-heavy, AP automation example: task/state/goal/tool-oriented content); inbound from Fortune 500.
• ICP: IT/Procurement Manager (discovers + implements), Finance/Ops Manager (identifies problem + tests), CXO of ~250-person company (approves + champions).

• 10x ARR growth in 18 months at 0 burn (period: ~May 2020 – Aug 2021 per chart axis).
• ARR chart shows continuous upward curve May 2020 → Aug 2021; no absolute ARR $ figure disclosed.
• NRR: 132%.
• Churn: <X% (specific value redacted).
• Free trial conversion: XX% (redacted).
• Traffic grew 3.7x in 12 months (Jul 2022 → Jul 2023).
• 34% of Fortune 500 have used the product.
• ARR projection chart (Series B deck): shows actual ARR bars Jan 2022 – Oct 2023, then projected bars to Jan 2025; absolute $ values not labeled on y-axis.
• Gross margin: >XX% (exact redacted).

• Gross margin: >XX% (redacted but stated as positive and high).
• NRR 132% implies negative net revenue churn - customers expand faster than they churn.
• Burn: 0 burn during 10x ARR growth period (base deck, ~2021). Burn status at Series B raise not stated.

• Competitive frame: Nanonets vs. Traditional OCR vs. AP-only tools:
• Traditional OCR: IT-friendly but needs developer setup; no learning; rule-based.
• AP-only tools (e.g., legacy AP automation): rule-based decisions; limited to one document type.
• Nanonets: end-to-end workflows; self-learning ML; processes any document format from day 0.
• ML moat: accuracy improves with volume (virtuous data flywheel); rule-based systems degrade with new formats.
• Time-to-live: ML path = days (AI training + ready); rule-based path = weeks (template creation + engineering + testing).
• Deep learning fine-tuning per customer creates switching cost: customer model is proprietary to their data.
• Named competitors: not explicitly named in deck. Implicit: UiPath, Automation Anywhere (RPA), ABBYY, Rossum, Hypatos (IDP).
• SaaS vs. in-house: large enterprise in-house build takes 6–12 months to production; SaaS wins on accuracy, speed, lower risk.

• CEO: Sarthak Jain; CTO: Prathamesh Juvatkar.
• 13 years building ML together (universities, startups, industry).
• EE + CS, IIT-GN 2012; previously co-founders at Cubeit.
• Team claim: "combined 100 years in Deep Learning and AI".
• Funding raised to date: $11.5M ($1.5M Seed YC Apr 2017 + $10M Series A Elevation Capital Oct 2021).
• Angels: founders of BrowserStack, Chargebee, Whatfix, PubMatic, Ally.

• Archetype + why: Usage-based SaaS ARR model with land-and-expand expansion revenue. The core billing unit is per-extraction (pages processed), with layered annual subscription and per-step workflow fees. NRR of 132% and a clear upsell motion from extraction → automation make this a classic land-and-expand PLG SaaS - cohort-based ARR tracking is the right frame. Not a pure seat-based SaaS; consumption volume and workflow depth are the key expansion levers.

• Forecast horizon & granularity: Monthly for Year 1–2 (given PLG dynamics and rapid ARR growth), quarterly for Years 3–5. 5-year total horizon to capture the S-curve from SMB PLG → enterprise expansion.

• Key drivers & assumptions:

*Revenue drivers:*

• Starting ARR: ~$2–4M ARR at Series B raise (~mid-2023), inferred from 10x growth in 18 months from a small base in 2020–21, then further growth through 2022–23. No absolute number in deck.
• New customer additions per month: split between self-serve (freemium → paid) and enterprise inbound; ratio ~70/30 at base.
• Average initial ACV (land): ~$500–2,000/year for SMB self-serve (based on $0.1/page × est. 500–2,000 pages/month); enterprise land contracts $10–30k/year.
• Pricing - extraction: $0.1/page entry rate; volume tier ~$0.0x/page; workflow step fee ~$0.0x/step.
• NRR: 132% - drive expansion via page volume growth + workflow step adoption + data type expansion ($xk/year add-ons).
• Gross churn rate: ~5–8%/year (implied by NRR 132% with net churn deeply negative; <X% monthly churn in deck).
• Free trial conversion: ~10–20% (XX% redacted in deck); model as a sensitivity variable.
• Enterprise pipeline: Fortune 500 penetration at 34%; upside from account expansion.

*Cost drivers:*

• Gross margin: ~70–80% (>XX% stated in deck; typical for ML API SaaS with GPU/inference costs; exact % unknown).
• COGS: GPU/inference compute, hosting (AWS/GCP), data labeling for model fine-tuning.
• S&M: ~25–35% of revenue at current stage; PLG model implies lower blended CAC vs. pure enterprise.
• R&D: ~30–40% of revenue (ML-heavy product; deep learning infra).
• G&A: ~10–15% of revenue.
• Burn: stated as 0 burn during 10x growth period; post-Series-B likely investment phase - model burn as S&M + R&D ramp.

*Operational drivers:*

• Pages processed per customer per month: seed value 500 SMB / 5,000 enterprise; grows as automation expands.
• Workflow steps per customer: starts at 0 (extraction only), grows to avg 3–5 steps/workflow as customers expand.
• Traffic growth as PLG input: 3.7x in 12 months; top-of-funnel conversion rate ~2–4% visitor to trial.

• Scenarios (Base / Bull / Bear - which variables flex):
• Base: NRR holds at 130–135%; new logo adds grow at market content-led rate; gross margin 75%.
• Bull: Enterprise penetration accelerates (Fortune 500 → $50k+ ACV); NRR expands to 140%+ as workflow automation fees compound; gross margin improves to 80%+ on scale.
• Bear: Competition from UiPath/Microsoft Power Automate compresses pricing; free-to-paid conversion drops; NRR falls to ~110%; gross margin pressure from increased GPU costs.
• Key flex variables: NRR, free trial conversion rate, enterprise vs. SMB mix, avg pages/customer/month, workflow step adoption rate.

• Required sheets / outputs:

1. Assumptions - all drivers in one place (pricing, volume, NRR, margins, headcount).
2. ARR Bridge - new ARR, expansion ARR (volume + workflow + data types), churned ARR, net new ARR; monthly then quarterly.
3. Cohort table - monthly cohort ARR with expansion curve; used to validate NRR.
4. Revenue schedule - extraction revenue (pages × price), subscription revenue (data type add-ons), workflow revenue (steps × price).
5. P&L - Revenue, COGS, Gross Profit, S&M, R&D, G&A, EBITDA, Net Income.
6. Cash flow - operating cash, capex (minimal), ending cash / runway.
7. KPI dashboard - ARR, MRR, NRR, gross margin %, pages processed, active customers, ACV by segment.
8. Scenario toggle - Base / Bull / Bear switcher on key assumptions.