A source-cited petroleum-geology knowledge base, an OverCaml-driven prospectivity screening layer, and a 3D subsurface explorer. Built entirely on public literature and open data to support Aramco's first-pass screening decisions.
Each point below is drawn from published public sources on Aramco's real programs. We access no proprietary data.
One of Earth's most underexplored basins, chasing Qusaiba-sourced unconventional potential (14 new fields announced Apr 2025).
aapg.orgDesigning fracs and sourcing/recycling water for a $100B unconventional program in an arid basin: seawater fracturing and produced-water recycling.
jpt.spe.orgA 9-Mtpa CCS hub injecting into a saline aquifer needs robust plume monitoring and MRV.
ccushub.ogci.comManaging the advancing water front and conformance across a peripheral seawater flood is a continuous optimization problem.
jpt.spe.orgNeeds far more compute for high-res depth imaging at scale (bought 7x more upstream capacity). We are not a competitor to GeoDRIVE.
aramco.comScaling generative AI (Metabrain) for well placement and seismic pattern detection on proprietary data. We do not position against Metabrain or EXPEC ARC.
offshore-technology.com90+ years and 7T data points under PDPL and NCA CCC-2:2024 in-Kingdom residency for critical national infrastructure.
nca.gov.saA complementary intelligence layer that feeds explorationists source-cited hypotheses. It does not replace their internal systems.
knowledge chunks from public petroleum-geology literature, with a source citation on every answer.
Multi-factor consensus scoring that ranks frontier hypotheses in confidence tiers. Literature-derived hypotheses, never reserves.
Interactive subsurface visualization for visual triage, with a zoom-triggered command-line interface.
Terminology, field and basin naming aligned to Aramco's working language and conventions.
Honest, concrete comparisons for the top pain points. Outputs are screening hypotheses your teams validate against your own data.
Scattered manual scanning of petroleum-systems literature across dispersed papers before prioritizing play areas in a vast, underexplored basin.
Literature-derived, confidence-scored play-fairway triage with explicit source citations, narrowing first-pass filtering to a shortlist geologists validate against their own seismic.
Screening frac and water-treatment technologies in an arid, analog-poor basin through fragmented searching of completion and water-management literature.
Retrieval-augmented search over tight-gas and water-treatment literature that shortlists frac-design and produced-water-recycling techniques with sources, ready for engineer review.
Manually assembling MRV protocols and monitoring-well design from scattered literature for a saline-aquifer CCS project.
A 3D-visualization plus RAG layer for CO2-plume methods, monitoring-well design and MRV-protocol scanning as pilot decision-support. It does not replace reservoir modeling.
Internal models trained on proprietary data are powerful, but cannot cite external peer-reviewed method provenance to justify an output.
A literature-grounded, citable second opinion with a transparent provenance layer that complements internal models rather than competing with them.
A simple, auditable pipeline. No seismic imaging, no reservoir simulation. Complementary to your stack.
Published petroleum geology and open data only. Text is not redistributed, only embeddings are stored.
bge-small embeddings with sqlite-vec vector search retrieve the most relevant chunks with their sources.
Multi-factor scoring ranks prospectivity hypotheses into confidence tiers. It shortlists, it does not decide.
3D visual outputs with a source citation on every result, for human validation.
Ready for in-Kingdom hosting to meet PDPL and NCA controls. Ingests only public or non-proprietary corpora unless deployed inside Aramco's environment.
The Jafurah and gas expansion raises demand for fast, cited screening of techniques and plays.
The Jubail hub and net-zero path need cited decision-support on monitoring and MRV methods.
A clear path to an in-Kingdom entity and a 5-year localization plan that raises award probability.
A vast, underexplored basin makes public-literature screening highest-value and lowest data-risk.
Soft entry first via Wa'ed Ventures or Aramco Digital, then the approved-vendor path.
Incubation via Wa'ed Ventures ($500M VC) or Aramco Digital, a proof-of-concept on public data.
Proven value on Rub' al Khali and Jafurah, with an in-Kingdom entity and cybersecurity certification.
A path to the Approved Vendor List, then acquisition that localizes the platform inside Aramco.
Start a scoped pilot on Rub' al Khali or Jafurah, entirely on public data. Prove the value as a complementary screening layer, then let us move toward acquisition and localization.
Soft entry via: Wa'ed Ventures · Aramco Digital SRM · srm@aramcodigital.com