QuickField.ai

Exploration-to-appraisal workbench · Oil & gas

Turn sparse discovery data into defensible production ranges — in minutes.

QuickField is a data-driven workbench for evaluating discoveries when subsurface data are still limited and a good development analog may not exist. Machine learning builds the type-well. You still choose the drilling plan, the economics.

Start after discovery. Refine the same assessment through appraisal.

QuickField total production view with ML, P10, P50 and P90 cases

The exploration problem

When the analog you need doesn't exist

Early field decisions still need quantitative answers. Traditional analog methods become weakest precisely where frontier and data-sparse discoveries create the most uncertainty.

How much might actually be recovered?
Which drilling plan makes development economic?
How wide is the uncertainty range?
Which appraisal data would reduce uncertainty most?

Validated against known outcomes

We tested the method blind

Historical reservoirs were held out one at a time. QuickField predicted production behaviour and recovery using the remaining data, then the predictions were compared with known outcomes.

73.0% 0 100
10-year cumulative oil · median accuracy
87.1% 0 100
Water-cut curve · median accuracy
87.9% 0 100
Recovery factor · median accuracy
What if the geology you are planning is not covered by our data?
Even if your geology is not in our source data, the approach still applies. We tested this by dropping the nearest 40% of analogs: ML remained more accurate than a conventional analog workflow.
Recovery factor — QuickField ML13.5%
Recovery factor — conventional analogs21.5%
Median absolute relative error after removing the nearest 40% of analogs. Lower is better.

One workbench

From discovery through appraisal

QuickField is the working environment around the decision, not a single AI prediction. The same asset evaluation becomes richer as appraisal progresses — from petrophysics and a type-well through drilling schedule, cash flow, and NPV.

  1. 01

    Start with what you have

    Petrophysics, PVT, reservoir assumptions and an early well-entry concept. Sparse data is expected.

  2. 02

    Build a virtual analog

    Use historical reservoir behaviour and ML instead of relying on one or two manually selected comparables.

  3. 03

    Generate production behaviour

    Estimate recovery factor, well production, oil and liquid rates, and water-cut evolution.

  4. 04

    Quantify uncertainty

    Run ranges and Monte Carlo to produce P10/P50/P90 outcomes and sensitivity to assumptions.

  5. 05

    Test development scenarios

    Change well count, timing and reservoir assumptions and compare alternatives in seconds.

  6. 06

    Add economics

    Connect subsurface uncertainty to cash flow, NPV and early investment decisions.

  7. 07

    Refine with appraisal

    Update the same assessment as new wells and better reservoir information become available.

  8. ROADMAP

    Expand the workbench

    Add capabilities around the exploration-geologist workflow without changing the core mental model.

Data foundation

Built on open data and public reports

QuickField is trained on open databases and public reports from oil and gas fields worldwide. For every field in our database you can inspect cumulative production history and the well operating model used to build analogs. Independents can start there. Operators can add company data behind their own tenancy.

Basin-level petrophysical ranges from open databases

Open databases

Basin and province coverage drawn from open geological and production databases — the foundation for matching analogs at regional scale.

Field map coverage from public oil and gas reports

Public reports

Field inventories and production records compiled from public reports, so every analog is grounded in documented historical performance.

Cumulative production history and base well operating model for a field

Per-field detail

For each field in the database you can review cumulative production and the well operating model — the same building blocks used to construct your analog.

What you get

The workbench, from catalog to economics

QuickField brings historical data, templates, drilling plans, ML forecasts and cash flow into one screening environment — so a greenfield case can be built, revised and compared in minutes.

Public production catalog

Thousands of fields worldwide, with production history and well operating models you can inspect and use as analogs.

Petrophysical atlas

Basin and province property ranges on the map — context for analog search and for setting layer attributes.

Company data vault

A protected store for your own fields and production history, isolated in the company tenancy and available as private analogs.

Petrophysical and economic templates

Reusable layer and cost presets so multi-iteration screening uses the same standards instead of re-entering assumptions each time.

Drilling schedule planner

Producers and injectors by year, start dates and injection ratio. Change the plan and production, water-cut and NPV update with it.

Well, layer and field estimates

Oil production, water-cut and recovery factor at well, reservoir and field scale, with checks that let you validate the ML calculations.

Monte Carlo production ranges

Sample reservoir assumptions to get P10, P50 and P90 profiles and see which inputs move recovery and rates.

Economic modelling

CAPEX, OPEX, oil price, injection, royalty, tax, discount rate and abandonment — NPV and cash flow on the same case, in tonnes or barrels.

Export

Download calculated results in standard formats (CSV / ZIP) for spreadsheets, data rooms and internal tools.

Who it is for

Same technology. Different reason to engage.

For independent E&P teams

Screen a discovery on our catalog from day one.

  • Get a first production envelope without assembling analog decks
  • Compare well counts until the case looks economic
  • Walk into a farm-out or partner meeting with P10/P50/P90 in hand
  • Revisit the same case as appraisal wells come in

Public fields and the petrophysical atlas are ready out of the box.

Open the workbench

For operators and majors

Run the same screening on your own production history.

  • Benchmark a new discovery against company fields, not only the public set
  • Test development concepts against corporate cost and reservoir templates
  • Give new-ventures and RE teams one quantitative case to argue from
  • Keep private analogs inside the company tenancy

Load your production history so the forecast sits on data you already trust.

Open the workbench

For exploration consultants

Evaluate more cases without rebuilding the analysis each time.

  • Independent discovery assessments
  • Farm-in and acquisition studies
  • Portfolio screening
  • Second opinions on operator forecasts
  • Client-ready quantitative benchmarks

Test QuickField on a historical client-style case before using it live.

Benchmark a client case

Two ways to build the type-well

Machine learning is the core. Manual mode is there when you want the wheel.

The value of QuickField is a data-driven type-well: models trained on historical reservoirs, scaled with reservoir-engineering calculations, then sampled with Monte Carlo. That is the default path from sparse petrophysics to a production range. If you prefer to assemble the curve yourself, manual analog mode keeps the same drilling schedule and economics — with inspectable spaghetti plots instead of a model score.

Core value

ML forecast

Use machine learning where historical data carry useful predictive signal. The model constructs a virtual analog from reservoir properties, then engineering relationships turn that well curve into a field forecast under your drilling plan.

  • Type-well oil, liquid and water-cut from layer attributes
  • Monte Carlo P10 / P50 / P90 and tornado sensitivity
  • Assumptions remain editable; uncertainty stays visible

Optional control

Manual analog mode

When you want to steer the type-well yourself, pick historical analogs and aggregate the curves. The drilling schedule and economics stay the same as in ML mode.

  • Search by field name or petrophysical ranges
  • Keep up to ten analogs per layer and inspect spaghetti plots
  • Aggregate mean or P10/P50/P90 year by year from the set you chose

A fair first test

See how QuickField reads a field you already know

Choose a historical discovery. Share only the information that would have been available at the time, and keep later production history to yourself. We run the assessment, then you compare it with what actually happened — or with your own estimate.

No polished demo case. Your geology. Your benchmark.

Independents can start on the public catalog. Operators can bring a proprietary analog set. Consultants can use a past client-style case.

Get in touch

Contact us for a demo

Tell us about the asset and where you are in the appraisal process — we will show you what QuickField makes of it.

From an estimated geological resource to a probability distribution of extractable resources — then a drilling plan and NPV you can defend.

Book a demo

Already a user? Open the application →