Breaking
NOW - Atlas AI maps emerging-market growth from low Earth orbit SERIES A - $7M led by Airbus Ventures, with Rockefeller + Micron STANFORD - Founded by Profs. Burke, Ermon & Lobell PRODUCT - Aperture Pulse detects change at planetary scale PBC - A Public Benefit Corporation by charter, not slogan STACK - Satellites + foundation models + ground truth
Company Profile - Geospatial AI

Atlas AI.
A planet, observed.

The Palo Alto company building the middleware between satellites and capital - and quietly teaching machines to read the world's economy from 400 kilometers up.

HQ Palo Alto, CA Founded 2018 Stage Series A Entity Public Benefit Corp
Atlas AI logo
The logo. Set against navy, the way a satellite sees a city at night. Atlas AI, Palo Alto, 2026.
The Dispatch

A view from orbit, a job on the ground.

Somewhere over the Sahel right now, a satellite is taking a picture. A few seconds later, another one. Then another. The images themselves are not particularly dramatic - tiled brown, the occasional green smear, a thin grey line where a road is being graded. To the human eye, mostly nothing. To Atlas AI, a signal.

Atlas AI does not own the satellites. It does not, strictly speaking, sell the images. What it sells is the answer to a question almost no one has been able to ask cleanly before: where is the world's economy actually changing, this month, by how much, and with what consequence?

The company calls itself a GeoAI platform. The phrase is dry. The work is not. Atlas AI's machine learning models stitch terabytes of satellite imagery to ground-truth surveys - household consumption data, agricultural yields, infrastructure inventories - and produce maps that look less like maps and more like dashboards for the built world.

"Monitoring drivers of economic development across the emerging markets so that financial capital can advance societal well-being."

That sentence is on the About page. It is also, conveniently, the legal mission of the company - because Atlas AI is a Public Benefit Corporation, an entity that promises in its charter to care about more than the cap table. Whether you find that earnest or strategic is, perhaps, beside the point. The product does the talking.

By the numbers

The shape of the thing.

2018
Founded
$7.2M
Total raised
~34
Team size
3
Stanford founders
The Product

Four things in one box.

Flagship

Aperture Pulse

Wide-area, near-real-time detection of economic and infrastructure changes - a heartbeat monitor for the built world.

Models

Geospatial Foundation Models

Planetary-aware ML pre-trained on satellite imagery and tuned with field surveys. Forecasts demand, growth, and risk.

Data

Analysis-Ready Library

Pre-processed datasets and a model catalog that drop into existing enterprise data science workflows.

Layer

Decision Middleware

Software that lets institutions monitor change, forecast impact, and simulate scenarios before deploying capital.

Use

Where it lands

Energy. Agriculture. Logistics. Real estate. Sovereigns and development banks. Anywhere a decision needs to know what the ground looks like next quarter.

Form

How you buy it

B2B subscriptions to data products and platform access. Engineering-friendly. API-first. Sold by humans who can spell "remote sensing."

The People

Three professors, one CEO, a planet's worth of data.

Marshall Burke

Co-founder - Stanford

Economist and Earth scientist. His research helped show that satellite imagery, paired with deep learning, could predict household consumption in places where census data lags by years.

Stefano Ermon

Co-founder - Stanford CS

Machine learning researcher. Brought the foundation-model thinking to Earth observation before the phrase "foundation model" became unavoidable.

David Lobell

Co-founder - Stanford

Director of Stanford's Center on Food Security and the Environment. The "what does this mean on the ground" voice in the founding trio.

Abe Tarapani

Chief Executive Officer

Operator-CEO. Runs the company day-to-day, hires globally, and writes the LinkedIn posts that turn quiet quarters into recruiting funnels.

Cap Table

An unusual set of backers.

Aerospace money. Semiconductor money. Foundation money. The Series A reads like a Venn diagram of who needs Earth observation next.

Airbus Ventures
Lead
Micron Ventures
Co-investor
Rockefeller Fdn.
Co-investor
Series A total
$7.0M

Source: Crunchbase, public announcements. Bar widths illustrative.

Make Earth's economic and social patterns observable, predictable, and actionable.
- Atlas AI, on what GeoAI is for
In Practice

What you can actually do with it.

Energy

Site a grid where the load is coming.

Forecast where demand will materialize before the spreadsheet does, using settlement and economic-activity signals derived from imagery.

Agri

Watch a yield, not wait for a report.

Combine remote sensing and ground truth to estimate crop performance across geographies that traditional surveys reach late or not at all.

Logistics

Stress-test a supply chain from space.

Use change detection to see infrastructure build-outs, congestion, and risk - before they show up in an ERP.

Capital

Underwrite emerging markets with new evidence.

Replace stale official statistics with high-resolution, frequently updated indicators of socioeconomic change.

A Brief Chronology

How we got here.

2018
Burke, Ermon, and Lobell incorporate Atlas AI as a Public Benefit Corporation, commercializing a decade of Stanford research linking imagery to socioeconomic measurement.
April 2020
$7M Series A led by Airbus Ventures, with Micron Ventures and The Rockefeller Foundation. An aerospace-semis-philanthropy triangle.
2024
Publishes "Mapping the Future," framing the company's thesis on planetary-scale economic intelligence. CEO Abe Tarapani opens hiring across Europe.
2026
Aperture Pulse and the geospatial foundation model library sit at the center of the product story. The middleware analogy is now load-bearing.
Watch & Listen

Interviews & demos.

Side Notes

Three things worth knowing.

Origin

Papers before pitch decks.

The founding team's academic work showed off-the-shelf satellite imagery + deep learning could predict household wealth in African villages better than some survey methods. The company is, in a sense, a paper that scaled.

Entity

The PBC is the point.

Public Benefit Corporation is a legal structure, not a marketing flourish. It binds the company's mission into its filings.

Naming

"Aperture Pulse."

Aperture as in the satellite optic. Pulse as in the heartbeat of an economy. Two precise words, doing the work of a paragraph.

Where To Find Them

Links, cleanly.

Share this profile
Closing

Back to the Sahel.

That satellite over the Sahel is still taking pictures. Somewhere in a server, a model is comparing this week's frame to last month's, last quarter's, last year's. A new road has been graded; a settlement has thickened; a stretch of solar panels has appeared where there was, on Tuesday, only dust.

Without Atlas AI, those frames are wallpaper. With it, they are an indicator - the kind of indicator a development bank can underwrite, a manufacturer can route around, an agency can plan against. The image has not changed. The question we can ask of it has.

Atlas AI P.B.C. - Palo Alto, California - established 2018.