This document details every roadmap item in the PetroBench platform architecture. Each item is organized by the layer it belongs to, with a description of what it does, the capabilities it delivers, and the business rationale for building it. Items marked In Development are partially built and moving toward general availability. All other items are planned.
Analytics, optimization, and design AI products. Draws on data from the Pulse Group (Well Monitoring) and RodSim to automate, recommend, and optimize. The five capabilities are designed to be built in the priority order below, with each one providing the foundation the next requires.
| Capability | What It Does | Business Value |
|---|---|---|
| Pulse-RodSim | Connects live well data to the simulation engine to close the surveillance-to-design loop | Reduces intervention lag, improves production uptime |
| API & MCP Layer | Exposes platform capabilities to external tools and AI agents programmatically | Makes PetroBench connectable to existing operator infrastructure and future AI tooling |
| Automated Design | Generates complete rod string designs from well parameters and targets | Scales design capacity without adding headcount |
| Assisted Design | Surfaces recommendations and flags issues during the design process | Faster time to design, consistent output quality across experience levels |
| AI Agent | Aggregates all intelligence capabilities into a single conversational and command-driven interface | Changes the product category from simulation tool to AI-powered engineering system |
Pulse-RodSim connects the Well Monitoring live data stream to the RodSim simulation engine. When a well deviates from its design targets, the platform analyzes the deviation, classifies the likely cause, generates a design recommendation, and routes it to the responsible engineer. Over time, with appropriate authorization controls, the loop can operate with increasing autonomy.
| Wellbore Grade | Assigns a continuous performance score to each well based on how closely actual operating conditions match the design targets. Grade changes trigger downstream analysis automatically. |
| Root Cause AI | Classifies the likely cause of a grade change or deviation using pattern recognition across well data, historical failure records, and production trends. |
| Pulse-RodSim | The live data bridge between Well Monitoring and RodSim. Supports three deployment patterns: side-by-side with an existing surveillance system (XSPOC, ForeSite, Ambynt) via API or database integration; replacement scenarios where the operator has an existing historian or SCADA database PetroBench pulls from; and direct-to-POC for greenfield deployments using Pulse POC hardware or a dedicated communication layer. The appropriate pattern is confirmed per operator before implementation scope is defined. |
| Sim. Refinement | Generates an updated simulation using current well conditions as inputs, producing a revised design recommendation that reflects what the well is doing now rather than at time of original design. |
The gap between surveillance and design in every operator's current workflow is measured in days to weeks. Pulse-RodSim compresses that cycle to hours and makes it systematic rather than reactive. The production uplift from faster intervention across a large well count is the most direct revenue argument in this roadmap.
The API and MCP Layer exposes PetroBench capabilities programmatically so that operators, third-party tools, and external AI agents can interact with the platform without the web interface. The platform UI leads by default, keeping IT involvement minimal and sales cycles shorter. The API layer gives companies the option to leverage the functionality how they want once they are already inside the platform.
| Sim. API | A versioned REST endpoint for submitting simulation inputs, retrieving results, and accessing historical simulation data. Enables operators to trigger and retrieve designs from their own data pipelines. |
| Well API | Programmatic read and write access to well data, including configurations, directional surveys, and production history. Required for operators wanting to keep PetroBench synchronized with upstream data systems. |
| MCP Server | An MCP-compliant server that exposes simulation, well data, and recommendation capabilities as callable tools for external AI agents. Positions PetroBench as infrastructure within the emerging AI agent ecosystem. |
| Dev Portal | Developer-facing documentation and sandbox environment. API key management, rate limit visibility, interactive API explorer, and usage analytics. Required for any operator or integration partner evaluating the API independently. |
Enterprise operators have existing data infrastructure: Snowflake, SCADA systems, production databases, and their own internal tooling. The API layer makes PetroBench connectable without requiring IT to get involved upfront. The MCP server makes PetroBench callable by any AI agent that supports the protocol, which is the infrastructure position that creates long-term platform defensibility.
Automated Design is a multi-run optimization pipeline that finds the best rod pump design for a well. The engineer defines evaluation criteria, equipment options, and a target production rate. The system generates hundreds of thousands of possible rod string and pump combinations, eliminates non-viable ones through progressive simulation runs, and returns the top designs ranked by a composite performance score. The engineer reviews the ranked results and applies the winning design directly to the well.
| Optim. Pipeline | Engineer defines evaluation criteria, equipment options, and target production. The system generates and evaluates hundreds of thousands of rod string and pump combinations automatically without manual simulation runs. |
| Prog. Elimination | Scenarios are filtered through a multi-stage funnel: static analysis removes geometrically infeasible combinations, SPM checks remove operationally unsound ones, and full simulation runs eliminate designs that fail against threshold criteria. Only the top-performing designs survive. |
| Ranked Results | Surviving designs are scored by a composite performance metric and ranked. The engineer reviews the results and applies the winning design directly to the well, creating a linked simulation record. |
| Run History | Every run is preserved per well. Previous runs can be cloned and modified to explore different constraint sets or equipment configurations without starting from scratch. |
Testing and iteration through end of April 2026. Demo and campaign mode from May onward. Commercial product scope to be defined once testing validates output quality.
Finding the optimal rod pump design for a well today requires running dozens of simulations by hand. Automated Design explores hundreds of thousands of combinations in a single run, surfacing configurations an engineer would never have time to find manually. It compresses design time from days to hours and improves output quality regardless of engineer experience level.
Assisted Design keeps the engineer in full control while the platform surfaces recommendations and flags issues that would otherwise require additional experience or manual analysis. The result is faster, more consistent design work across teams of varying seniority.
| Recommendations | The platform analyzes well parameters and proposes optimal starting configurations for rod string design, pumping unit selection, and pump depth, reducing blank-page time for each new design. |
| Feedback | As the engineer builds a design, the platform evaluates each configuration change in real time and flags combinations that fall outside acceptable operating bounds before the simulation runs. |
| Config Review | A structured review pass that checks the completed design against known failure patterns, API standards, and the operator's own historical performance data before the engineer commits. |
| Auto-Rod Table | Automatically generates a rod taper table from the well geometry, target load, and available rod grades, giving engineers a validated starting point rather than building from scratch. |
The median rod lift engineer manages 300 to 500 wells. Assisted Design removes the repetitive pattern-matching work without removing the engineer from the decision. For operators evaluating PetroBench against incumbent tools, this is a visible productivity argument that affects headcount planning and onboarding time for junior engineers.
The PetroBench AI Agent aggregates all intelligence layer capabilities into a single interface. Engineers and operators interact with design recommendations, live optimization outputs, portfolio analysis, and reporting through a conversational interface, direct commands, and a consolidated dashboard.
| Chat Bot | A conversational interface through which engineers can query well performance, request design recommendations, review optimization suggestions, and get answers about specific wells without navigating the full application. |
| Auto Report | Generates design review summaries, intervention logs, and performance trend reports on a scheduled or event-triggered basis. Delivers outputs to the engineer's inbox or into connected reporting systems. |
| Commands | Structured command syntax for engineers who prefer direct instructions over conversation. Commands can trigger simulations, request batch analyses, approve or reject pending designs, and surface specific well data. |
| Consolidated UI | A single dashboard surfacing the most relevant information across all intelligence layer outputs: wells with grade changes, pending approvals, active recommendations, and auto-generated reports in one view. |
The AI Agent changes the category PetroBench competes in. It is no longer a simulation tool or a surveillance platform. It is an AI-powered engineering system for rod lift operations. That framing commands a different pricing conversation, a different buyer, and a different competitive landscape.
PB Aware extends RodSim's wave equation solver to legacy rod pump controllers that produce deviated or non-standard dynamometer card formats. It corrects for the signal distortions introduced by older POC hardware so that engineers can run accurate downhole card reconstruction on wells they currently cannot analyze in RodSim without manual preprocessing.
| Deviated Wave Eq. | Applies a correction layer to incoming surface card data from legacy POCs where the raw signal does not conform to the standard input format expected by the RodSim wave equation solver. Produces a corrected surface card that the standard pipeline can process without modification. |
| Legacy POC Support | Extends RodSim compatibility to controller models that are currently excluded from automated analysis due to signal format limitations. The operator's existing POC hardware remains in place — no field equipment changes are required. |
| Card Quality Flag | Surfaces a confidence indicator on each corrected card so engineers know when the reconstruction is based on a corrected input versus a native-format signal. Supports engineering judgment on whether to act on the result or inspect the well directly. |
PB Aware is a diagnostic-first product. The primary sales trigger is an operator who has run a workover spike investigation, started a new artificial lift program, or attended a conference session on dynacard analysis and realizes their current POC hardware is excluding a portion of their well count from automated analysis. The production engineer identifies the gap; the operations manager approves via LOE or opex budget without a capital authorization.
Operators running mixed fleets of controller hardware have wells that fall outside standard analysis workflows. PB Aware converts those wells from manual review burdens into standard platform participants without a hardware refresh. For operators with 20% or more of their well count on legacy controllers, the value argument is direct and quantifiable.
The Well Monitoring product has two items live today: the Dashboard and Pulse POC. The three roadmap items below expand it from a view-only dashboard into an active surveillance and alerting system, and are prerequisites for Pulse-RodSim.
| Surveillance | Continuous monitoring of well operating parameters against design targets. Tracks pump fillage, surface card shape, stroke rate, and fluid levels relative to the original simulation outputs. |
| Failure Tracking | Logs and classifies well failures and anomalies over time. Builds a failure history per well that feeds pattern recognition in the Pulse-RodSim layer and informs preventive maintenance scheduling. |
| Live Alerts | Configurable alert rules that notify engineers when a well crosses defined thresholds. Supports email, in-app, and eventually API webhook delivery. Alerts are linked to the well record and tracked for response time. |
These three capabilities position Well Monitoring as a full surveillance layer in two distinct deployment scenarios. In displacement scenarios, Well Monitoring replaces XSPOC or similar platforms as the operator's primary surveillance tool. In side-by-side scenarios, Well Monitoring integrates with existing systems as a complementary layer, with PetroBench pulling data from the operator's current infrastructure rather than competing with it. Without surveillance, failure tracking, and alerts, the product surfaces historical data without driving action. With them, it becomes the system engineers rely on to know which wells need attention and why.
The Vendor and Reseller section has three live capabilities today: Vendor Portal, Product Catalog, and Custom Reports. The Reseller capability enables the Weatherford OEM channel and any future reseller relationships.
Creates a structured reseller tier within the platform that allows a named reseller to provision and manage client accounts under their own branded environment. The reseller has visibility into all accounts they have created, can manage license assignments, view usage data, and surface their own product catalog to clients. PetroBench retains the platform relationship and billing infrastructure.
The Weatherford OEM agreement is the primary driver. Without this capability, the reseller relationship requires manual account creation and management by the PetroBench team. The Reseller capability makes the channel self-sufficient, which is required for the relationship to scale beyond a handful of pilot accounts.
| Compliance | Generates structured compliance reports documenting design decisions, simulation inputs, approvals, and change history in a format suitable for regulatory review or internal audit. |
| Audit Log | A complete, exportable record of every user action, data change, simulation run, approval, and configuration modification. Searchable by user, date, well, and action type. Supports both internal governance and external audit requirements. |
Enterprise operators, particularly those in regulated environments or with internal change management requirements, need documented evidence that engineering decisions were reviewed, approved, and traceable. These two capabilities are often the difference between a platform being approved for enterprise deployment and being classified as an uncontrolled tool during IT and legal procurement reviews.
Analytics is a fully roadmap platform component that uses data accumulated across wells, simulations, designs, and interventions to surface patterns that inform preventive action, process improvement, and organizational standardization.
| Preventive Maint. | Identifies wells statistically likely to require intervention within a defined window based on failure history, grade trends, and operating condition patterns. Outputs a prioritized maintenance schedule rather than a reactive failure list. |
| Optimization | Surfaces portfolio-level design and operational patterns that correlate with better performance outcomes. Identifies which rod string configurations, pump types, or operating parameters consistently outperform alternatives across similar well conditions. |
| Standardization | Flags divergence from internal design standards and preferred configurations across engineering teams or regions. Supports the enforcement of division-level design templates and highlights where engineers deviate from established best practice. |
| Data Insights | A configurable analytics dashboard allowing operations and engineering managers to query portfolio performance, track KPIs over time, and export structured data for reporting into connected data lake infrastructure. |
Analytics converts PetroBench from a transactional tool into an organizational asset. A year of simulations, outcomes, and interventions produces institutional knowledge that the platform can surface, standardize, and act on. This is the capability that justifies the platform to engineering managers and VP-level buyers evaluating whether PetroBench improves the productivity of their entire engineering organization.
The balancing calculator is the next addition to the Calculators section once requirements are defined. It addresses a routine part of rod lift design work that engineers currently perform outside the platform in spreadsheets or separate tools.
Calculates the counterbalance required for a given pumping unit and rod load configuration. Takes rod string weight, fluid load, and pumping unit geometry as inputs and returns the required counterbalance position, moment, and recommended adjustment to minimize peak torque and extend equipment life.
Requirements gathering in progress. Build phase to begin once requirements are finalized. Not scheduled for EOQ2.
If engineers are running this calculation outside PetroBench, the platform is not yet their complete design environment. Adding it reduces context switching and reinforces platform completeness in the sales cycle.
ForeSite is Weatherford's well control and production optimization platform. PetroBench has an OEM reseller agreement with Weatherford in which PetroBench is positioned as the design engineering layer that complements ForeSite's control and monitoring capabilities. A formal integration requires a dedicated discussion with Weatherford before any development scope is defined.
Any integration must be framed as RodSim design outputs flowing into ForeSite, not as PetroBench competing with ForeSite's surveillance and control capabilities. The Well Monitoring product overlaps with ForeSite in the surveillance space. That distinction must be managed carefully in any technical or commercial conversation with Weatherford about integration scope.
| Design Export | Export RodSim design outputs in a format ForeSite can ingest directly, eliminating the manual step of re-entering design parameters into ForeSite after a simulation is completed. |
| Live Data Import | Surface ForeSite operational data within PetroBench for engineers running design updates against current well conditions, closing the gap between the two platforms for engineers who use both. |
| Joint GTM | Coordinate positioning for operators who run both platforms, presenting PetroBench and ForeSite as complementary layers in a complete rod lift engineering and operations stack. |
If Weatherford's existing operator base is already running ForeSite, a native integration reduces the friction of adding PetroBench to those accounts and strengthens the OEM relationship. Whether this is commercially viable depends on Weatherford's appetite and how the Well Monitoring competitive dynamic is resolved. Treat as conditional until that conversation produces a clear direction.
Internal use only. Prepared by PetroBench Business Team — April 2026. Not for external distribution.