New: Automated Design is live

Enterprise Integrations

Connect Your Stack to the Engine

Lake connectors, legacy import, identity, REST, and MCP, same credentials when AI is on the roadmap.

Databricks
WellView
AWS
MCP
SAML
Okta
Azure
Snowflake

Lakes, identity, and agent surfaces, same credentials and permissions

Data Lakes and Well Systems

Connectors for the Lakes. Import for Desktop Tools

Pull wells, surveys, and production from Snowflake, Databricks, Azure, AWS, and WellView into a simulation in one fetch. Bulk import covers common RodStar and SRod formats when wells are still on the desktop.

Snowflake
Databricks
Azure
AWS
WellView
More Structured lakes

Enterprise connectors

Snowflake, Databricks, Azure, AWS, and WellView. Pull into PetroBench well records under the same org scopes your engineers already work inside.

RSDX file RSDX RodStar design
RSVX file RSVX RodStar vertical
INP6E file INP6E SRod input
XLSX file XLSX Spreadsheet

Legacy bulk import

Bring RodStar and SRod designs into PetroBench without rebuilding well setups by hand. Common desktop formats land in the same workspace as lake-fed wells.

Identity SSO and SCIM

Your IdP Stays the Source of Truth

Engineers sign in with credentials IT already manages. SCIM provisions and deprovisions users, no second user directory to maintain.

Identity providers
  • Microsoft Entra ID Azure AD
  • Okta Workforce Identity
  • SAML / OIDC Any compliant IdP

Single sign-on

SAML 2.0 and OpenID Connect through the IdP you already run.

SCIM provisioning

New hires get accounts automatically. Deprovision revokes access.

No parallel directory

IT keeps one source of truth. PetroBench consumes it.

Surfaces One Credential, Three Paths

Mint Once. Call REST, MCP, or Vector

The same scoped Bearer token on every surface. Point your AI stack at MCP, your pipelines at REST, and keep engineer signoff on live-well writes.

Explore MCP
credential / surfaces org-scoped, Bearer
Authorization Bearer pb_svc_org_8f2a9c…
Scopes
  • wells:read
  • simulations:read
  • organization:read

Granted on this token

AI When the Company Runs Agents

Built for the AI Program Already Underway

If leadership is funding agents, PetroBench already exposes rod-lift wells over MCP and REST. IT keeps the scopes. Engineers keep signoff on writes.

MCP for the agent stack

Point Claude, Cursor, or your internal agents at app.petrobench.com/mcp. Same wells and RodSim tools under org scopes.

REST when you build your own

app.petrobench.com/api/v1 for pipelines and internal apps. One Bearer token, same permissions as the UI.

Vector in the product

In-product chat on the open well. Live-well writes stay behind engineer signoff and the audit log.

How IT Works

From Your Data to a Decision

Connect sources once. Run RodSim on the well record. Engineers review designs before anything writes back.

Your data
Snowflake
Databricks
WellView
AWS
Azure
Any source
PetroBench Platform
Design
RodSim
RodSim
Rod lift design
Dynamics
Dynamics
Dynamic analysis
Calculators
Calculators
Eng. tools
Automated Design
Automated Design
Constraint-driven design
Platform layer
Well Management
Unified well database
User Management
Role-based access
Company Mgmt
Corporate settings
Event Hub
Multi-Tenancy
Workflows
Data Repository
SSO
Security & Audit
Your agents
Claude
OpenAI
Gemini
Cursor
MCP
REST API
CLI

Connect Your Stack on One Pilot Well

Start with a well from your lake or a legacy file. Same credentials on REST, MCP, and Vector.