Turning the world’s science into decisions

Krigara is an AI that reads the scientific literature and the data, runs the models, and makes the call, with the evidence to back it. Proven first underground, built for any science.

Our goal

Make every scientific decision on all of the evidence.

The problem
Scientific articles published each year, worldwide, and our projection to 2030
01234562014: 2,193,498 articles20142015: 2,250,669 articles20152016: 2,335,510 articles20162017: 2,425,123 articles20172018: 2,522,582 articles20182019: 2,710,512 articles20192020: 2,873,342 articles20202021: 3,126,671 articles20212022: 3,238,106 articles20222023: 3,285,108 articles20232024: 3,525,866 articles20242025: about 3.80 million articles (projection)20252026: about 4.09 million articles (projection)20262027: about 4.40 million articles (projection)20272028: about 4.74 million articles (projection)20282029: about 5.11 million articles (projection)20292030: about 5.50 million articles (projection)2030 Year Articles (millions)
Published (NSF, Scopus)Projection to 5.5 million in 2030 (7.7% a year)
3.5 million
new scientific articles published in 2024, heading for 5.5 million a year by 2030
vs
264
articles a typical scientist reads in a year
Sources: NSF/NCSES, The State of U.S. Science and Engineering 2026, Figure 29 (Scopus, accessed August 2025); no 2025 count has been published yet, so 2025–2030 are a projection reaching 5.5 million in 2030, a growth of 7.7% a year (the 2014–2024 average was 4.9%). Reading: Tenopir et al. survey, reported in Nature (2014).

Every year, science publishes more than ten thousand times what any one scientist can read. Most of what is known never reaches the decision it could change.

We don’t lack knowledge. We lack the time to use it.

Our solution

Krigara combines AI and human expertise to make scientific decisions faster, and to show its working.

Knowledge

The literature, as data

1.55 million papers and more than 40 million measurements, each with its units and its source. See a real example

Models

It runs the science itself

The field’s own models and software, operated by the AI and checked against measurements.

Decisions

With evidence and uncertainty

A decision a board can act on, in hours instead of months, with every number traceable.

Learn about the science

Applications

One decision engine. A first application where a drill bit proves it, and a roadmap across science.

Live

Earth resources

Natural hydrogen, CO₂ storage, critical minerals and geothermal heat, in the ultramafic rocks where they occur together.

Tested

Plasma chemistry

Hydrogen and ammonia made with plasma: a reaction model repaired from the primary literature.

Chemistry and materials

Catalysts, reactions and materials, where the answers sit in thousands of published measurements.

Any data-rich science

Wherever the evidence already exists, scattered, and the decision is worth money.

Explore applications

Our first decision

A hydrogen target in an exploration well.

Krigara combined the literature with magnetic, gravity and seismic data to map natural hydrogen in ultramafic rock. At the heart of that map, an exploration well proved gas at around 2,000 metres; it will be re-entered in 2027 to test how much it can produce. Its second decision, a CO₂ storage site in Greece, has since won a €29M EU Innovation Fund grant.

See the applications
1.55 M
Scientific papers read and indexed
40 M+
Measurements extracted, with units and sources
782,000
Rock analyses in one structured database
Hours
From question to decision, instead of months
Proven with
DecahydronWeatherfordEU Innovation FundFOIZ

Let’s decide from the evidence.

Tell us the question, and the data you have. We will show you what Krigara does with it.

Contact us