# Kili Technology Financial Model

Training data management platform that helps enterprises label, manage, and improve data quality for AI/ML models.

- Canonical: https://finamodel.com/startups/kili-technology
- Excel download: https://finamodel.com/startup-models/kili-technology.xlsx
- Category: AI/ML
- Model type: SaaS ARR / Valuation
- Funding round: Series A
- Funding: $25M
- Founded: 2021
- Geography: Not in deck (company appears French-founded based on "PRÉDICTIONS" in slide 6; HQ not stated).
- Customer: B2B

## About the company

Kili Technology provides a data-labelling platform for machine-learning teams. Customers use it to organise training-data workflows, manage annotation quality, and coordinate the people and processes required to build reliable datasets for AI development.

The enterprise product can be sold as recurring platform access and expand as customers add users, projects, data volume, and model-development teams. Its economics must recognise that annotation and quality-control delivery can carry variable costs beyond conventional software hosting.

The model starts with enterprise contracts and a cohort-based ARR bridge for new logos, expansion, and churn. It layers users and labelled-data volume onto revenue and COGS, then tracks gross margin, delivery capacity, sales and R&D headcount, operating cash flow, 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

- Platform for managing training data end-to-end: labeling, collaboration, quality monitoring, and labeling automation.
- Supports all asset types (image, text, audio, video implied).
- Human-in-the-loop workflow: combines ML predictions with human annotation to accelerate labeling.
- Data pipeline integration and ML application output monitoring.
- Core promise: turn raw, unstructured data into "ground truth" at industrial scale.

## Market

- 80% of unstructured data by 2025 - $733bn market size (sourced from Markets & Markets, Grandview Research, GM Insights).
- 80% of enterprises investing in AI.
- 80% of AI projects not deployed in production.
- No SAM or SOM breakdown provided.
- No explicit data-labeling-software market size cited (the $733bn appears to reference the broader unstructured data / AI market, not the labeling tooling TAM specifically).

## Competition / moat

- Competitive context referenced indirectly: Google (annotation since 2013, "Bighead Architecture"), Facebook (user-scale annotation), Apple, Lyft, Amazon, Uber - all built proprietary annotation platforms.
- Positioning: Kili makes available to enterprises what tech giants built internally.
- No direct competitor names (Scale AI, Labelbox, Snorkel, etc.) mentioned.
- Moat claims: data quality impact quantified - 10% reduction in label accuracy typically results in 2–5% model performance decrease; up to 10% errors found in most-used open-source datasets.

## Recommended financial model

- **Archetype + why:** B2B SaaS ARR model. Kili sells a platform (seat- or usage-based license) to enterprise data/ML teams. The natural model is ARR-driven with expansion revenue as customers label more data types and add annotators. No transaction/marketplace element is described.
- **Forecast horizon & granularity:** 3 years monthly (M1–M36), collapsing to quarterly from Y2 for readability. Early-stage deck warrants monthly granularity to track cash burn and ARR ramp.
- **Key drivers & assumptions:**
  - New logos / month: ramp from 2 → 8 new customers/month over 3 years; no customer count in deck
  - ACV (Annual Contract Value): $60k–$120k enterprise ACV; comparable to Labelbox/Scale AI mid-market; not in deck
  - Revenue per customer: seat-based or data-volume-based; modelled as flat ACV with net revenue retention
  - Net Revenue Retention (NRR): 115–125% for B2B SaaS AI tooling with expansion from more seats/data types
  - Gross margin: 70–80%; platform software with some cloud compute cost; no COGS disclosed
  - CAC: $30k–$50k enterprise sales motion; no sales data in deck
  - Payback period: 12–18 months given assumed ACV and gross margin
  - Headcount: sales + CS hiring tied to new logo targets; no team data in deck
  - R&D spend: 40–50% of revenue at this stage (early SaaS)
  - Churn: 5–10% gross annual logo churn; platform stickiness from data lock-in supports low churn assumption
  - Market size: $733bn addressable market cited; data-labeling software sub-segment is meaningfully smaller - ~$3–5bn by 2025 for software tooling specifically
- **Scenarios (Base / Bull / Bear - which variables flex):**
  - Base: 4 new logos/month at $80k ACV, 120% NRR, 75% gross margin
  - Bull: 7 new logos/month, $100k ACV, 130% NRR (platform becomes standard for large enterprise ML teams)
  - Bear: 2 new logos/month, $60k ACV, 110% NRR, 68% gross margin (longer sales cycles, pricing pressure from Scale AI / open-source)
  - Key flex variables: new logo velocity, ACV, NRR, gross margin
- **Required sheets / outputs:**
  - Assumptions (all drivers, clearly flagged vs)
  - ARR Bridge (new ARR, expansion, churn, net new ARR)
  - P&L (Revenue, COGS, Gross Profit, OpEx by function: R&D, S&M, G&A, EBITDA)
  - Headcount plan (by department, tied to revenue drivers)
  - Cash & runway (burn rate, months of runway, fundraising trigger)
  - Scenario toggle (Base / Bull / Bear)
  - KPI dashboard (ARR, MRR, logos, NRR, CAC, LTV, gross margin, burn multiple)

## Frequently asked questions

### Is the Kili Technology financial model free?

Yes. The Kili Technology 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.