CLARISSA · DEMO

CLARISSA AI reservoir simulation demo

See the workflow CLARISSA is built around today—and the end-to-end research path being tested from sealed field records to ensemble forecast.

Use the demo request form to discuss a demonstration or beta access.

QR code to request a CLARISSA demonstration

SCAN FOR ACCESS

Scan the QR code or tap it to request a CLARISSA demonstration or beta access.

Reservoir simulation is the demonstrated capability.

01 · DESCRIBEState the reservoir task in engineering terms.
02 · BUILDGenerate a structured reservoir simulation model.
03 · VALIDATEApply deterministic checks and simulator-aware gates.
04 · RUN + ANALYZEExecute the model and inspect the engineering result.

What to bring to a demo conversation

Describe the reservoir engineering task you want to explore, the simulator workflow you use and the outputs you need to review. Start with a high-level description so the discussion can focus on model authoring, verification and simulator results.

For technical background, read how CLARISSA, PetroScript, OPM Flow and RIGOR fit together.

From sealed field records to an ensemble forecast.

The end-to-end harness is being tested on a scientific workflow in which CLARISSA reads a sealed evidence bundle, keeps observation, derivation, interpretation and assumption distinct, builds executable reservoir models, registers ensemble members before execution, and produces forecast bands from the resulting runs.

FIELD RECORDSDocuments, logs, surveys, spreadsheets and production reports.
EVIDENCEExtract facts and carry provenance forward.
INTERPRETForm hypotheses and preserve consequential alternatives.
MODELConstruct executable reservoir cases.
VERIFY + ENSEMBLERegister, run and replay complete members.
P10 · P50 · P90Publish probabilistic forecast bands.
Beta status: prompt tuning, verification gates and operating modes are still being refined through ATCE. This page describes the workflow being tested; it is not a claim that every raw-field-data case is production-ready today.

The capabilities expanding toward ATCE.

Reservoir-only mode + gate tuning

Refine the reservoir workflow and tune verification gates for external testing.

DWSIM + GEOS

Expand toward pipe-network and stronger compositional workflows.

Model selector + BYOK

Run the same harness with locally deployed models or a user-selected frontier model.

Voice mode

Add conversational voice interaction for engineering workflows.

Prompt + trajectory optimization

Use verified execution traces for GEPA-style and related optimization approaches.

Future fine-tuning

Build a governed corpus and reward infrastructure that can support later model-weight specialization.