How Pachama uses satellite data and AI to fund forest conservation

Executive overview

Carbon markets were designed to fund climate solutions, but less than 2% of that capital reaches forests — the most scalable carbon capture tool available. The barrier is verification: certifying a single forest project costs up to $400,000 and takes two years.

Pachama replaces physical audits with remote sensing and machine learning, cutting verification cost and time dramatically. It then connects certified forest projects directly to corporate carbon buyers through a transparent marketplace.

Forest conservation is one of the largest carbon capture opportunities on Earth — the bottleneck is data, not land.

The carbon market problem

• Companies offsetting emissions must buy carbon credits — issued to projects that reduce or capture CO2
• Less than 2% of ~$80B flowing through carbon markets annually reaches forest projects
• Certification requires physical audits: two years and up to $400K per project
• Fewer than 50 US forest projects have achieved carbon credit certification
• Past fraud and lack of transparency eroded buyer trust in forest offsets
• Most capital flows through brokers and traders, creating opacity and inefficiency

How Pachama's verification works

• Uses LiDAR point clouds (from drones, aircraft, satellites) and satellite imagery to measure forest biomass
• Convolutional neural networks trained on ground-truth field plots predict carbon stored in a forest
• Demonstrated under 1.5% error in predicting carbon stock across New England forests
• Models also detect biodiversity markers: canopy shape distinguishes conifers from broadleaf; tree density signals wildlife habitat
• NASA's GEDI LiDAR instrument (on the ISS) is releasing global forest data — Pachama plans to use it to extend models planet-wide
• Continuous satellite monitoring replaces periodic physical audits, enabling real-time project tracking

The marketplace

• Pachama lists verified forest projects on a platform with satellite imagery, carbon calculations, and co-benefit descriptions
• Buyers access projects directly, bypassing brokers and traders
• Co-benefits shown per project: biodiversity, water conservation, community impact, job creation
• Pricing: carbon credits currently range $5–$20/ton; offsetting one employee at a service company costs roughly $50–$150/year
• Over 10 projects onboarded in the US and Brazil; over 10 customers closed at time of recording

Four types of forest projects

1. Forest conservation — paying landowners not to clear standing forest (e.g. Amazon); preserves existing carbon stock
2. Improved forest management — sustainable timber harvesting with replanting commitments; credits issued on the carbon-positive delta
3. Reforestation — planting trees where forest previously existed
4. Afforestation — converting non-forest land (e.g. degraded desert) to forest; hardest and least urgent given available restoration land

Scale of the opportunity

• ~1 billion hectares globally available for reforestation without competing with agriculture
• Planting capacity: ~1 trillion trees, with potential to sequester ~200 gigatons of carbon (roughly two-thirds of post-Industrial Revolution emissions)
• Voluntary carbon market growing ~50% year-over-year since the Paris Agreement
• Airlines mandated to offset emissions from 2021 (Corsia framework): 160M tonnes/year of new demand
• Demand drivers: regulatory compliance, employee and investor pressure, consumer sustainability expectations

Who buys carbon credits

• Compliance buyers: energy, transport, and industrial companies facing regulatory caps
• Voluntary corporate buyers: tech companies, brands for whom sustainability is a value proposition
• Airlines: moving from voluntary to mandatory; Delta and United already offer passenger offset options
• Future: governments, regulated industries, corporates, and consumers all expected to participate

Building trust in forest offsets

• Previous credibility problems stemmed from paper-based tracking and inability to detect fraud or illegal deforestation in real time
• Pachama's approach: digital audit trail from issuance through trading, verifiable online
• Double-counting risk addressed through existing registries; more transparency tooling still needed
• National carbon accounting (country Paris commitments) must be reconciled with private project credits — a data standardisation problem

Fundraising and mission alignment

• Easiest fundraise of Diego's career; attracted top investors by framing climate as the largest value-creation opportunity
• Lead investor: Ryan Graves (Uber co-founder); others include Chris Sacca, Paul Graham, Social Capital, Global Founders Capital, Accel, Atomico
• Raised nearly twice the initial target
• Pitching approach: identical to any scalable tech startup — no special framing for climate; YC's Silicon Valley playbook applied directly
• Investors told upfront: this is a long-term, mission-driven company; some declined, most appreciated the transparency

Advice for founders and engineers entering climate tech

• Take 6–12 months on research before building: read, talk to researchers, professors, and executives, discard ideas
• Domain experts are open to conversations with motivated outsiders
• High-potential AI + X intersections: shipping optimisation, air travel, nuclear safety, smart cities, electric grid management, precision agriculture
• Within big companies: find the climate-focused team and contribute there
• Start a company, join a startup, or work within a large company's climate division — all valid paths