Exploration decisions still run on
spreadsheets and phone calls.
Independent operators evaluate prospects using inherited knowledge, personal networks, and manual data pulls. The result is slow, inconsistent, and impossible to validate after the fact.
Intelligence at every stage of evaluation
From public data to scored prospects
Start in your basin
ProspectOS has already ingested every public well, permit, production record, and government lease across the DJ Basin and Permian, with additional basins in adapter staging. No data upload required.
Prospects scored automatically
The engine clusters producing wells by formation and spatial proximity, detects coverage gaps, and generates ranked prospect polygons. Each prospect receives a score with a P10/P50/P90 range and its contributing factors decomposed.
Evaluate the acreage
Draw a polygon or select a section on the map and get a real-time score with its evidence chain — analog wells, geological signals, operator track record, and surface constraints. Ask the Exploration Advisor a question and get the same answer the map gives. Land, management, and investment team views are one click from the same data.
Track and validate
Set watch zones on areas of interest. As new permits are filed and wells are drilled, ProspectOS matches outcomes to predictions and updates your track record. The model improves with every drilling result.
Built different from the ground up
Calibrated, not just predictive
Every score is benchmarked against historical outcomes. We publish our hit rates.
Explainable by design
SHAP attribution on every score. Know exactly which features drive each prospect's ranking.
Public data, no lock-in
Built entirely on government regulatory data. No proprietary data dependencies or vendor lock-in.
Operator intelligence included
Quantified operator reliability — no other tool offers this from public production data.
Basin-agnostic by design
DJ Basin and Permian today, with basin-aware benchmarks as we expand. Coverage is a state-of-coverage statement, not a definition of who the product is for.
Built around your acreage
Draw a polygon and get a scored section with its full evidence chain in seconds. Evaluate the area, not a list someone else generated.
Verifiable track record
Every score is timestamped and matched to drilling outcomes. We publish the calibration backtest — honestly, including where the model is modest.
AI advisor built in
Ask any question about any location. The Exploration Advisor reasons across the full dataset.
Rigorous methodology, transparent results
Gradient Boosted Trees
A 35-feature gradient-boosted model with spatial block cross-validation and P10/P50/P90 quantile regression, validated against a transparent heuristic baseline — the better performer runs in production.
Spatial Feature Engineering
PostGIS-native feature extraction: analog well density, production quality within radius, formation-level benchmarks, and drilling recency signals.
SHAP Explainability
SHapley Additive exPlanations provide per-prospect, per-feature attribution scores. Every deviation from baseline is accounted for.
Prediction-Outcome Loop
Prospective validation framework matches timestamped predictions to real drilling outcomes. The model retrains on expanding ground truth.
Built on public regulatory data
Every well, permit, production record, and government lease in ProspectOS comes from state and federal regulatory agencies. No proprietary data dependencies, no vendor lock-in. Your analysis is reproducible and auditable.
Ready to evaluate prospects
with confidence?
See how ProspectOS scores prospects in your basin. No implementation, no data upload, no commitment.
Perspectives on upstream exploration
How AI Is Reshaping Oil and Gas Exploration
The upstream industry is experiencing a quiet revolution. Machine learning models trained on public well data can now score drilling prospects with measurable accuracy.
What Makes a Good Drilling Prospect?
A framework for evaluating upstream opportunities using production analogs, geological attributes, operator track record, and spacing analysis.
Multi-Basin Exploration for Independents
How small operators can use publicly available data to evaluate opportunities across basins without Enverus-scale budgets.