Transcend adds AI tools to its deterministic engineering platform

Transcend on Oct. 13 introduced two enterprise AI capabilities for its Transcend Design Generator: Ask TDG and MCP Gateway. The update is meant to help utilities, OEMs and engineering teams work faster with trusted preliminary design data without losing the deterministic engineering behind it.

Why it matters: - Utilities and infrastructure teams are under pressure to do more planning with limited engineering capacity. - Transcend’s new AI tools are meant to speed capital planning, project delivery and technical sales while keeping engineering outputs traceable and governed. - The release targets a core risk in enterprise AI: using AI around engineering data without letting a non-deterministic model alter the calculations.

What happened: - Transcend on Oct. 13 announced new enterprise AI capabilities for its Transcend Design Generator, or TDG. - The two features are Ask TDG and MCP Gateway. - Ask TDG is a grounded AI assistant inside TDG. - MCP Gateway lets authorized AI environments access selected TDG engineering capabilities through Model Context Protocol, an open standard. - The announcement was made in Princeton, New Jersey.

The details: - TDG is Transcend’s generative engineering platform for water, wastewater and power infrastructure. - The platform uses deterministic, traceable, first-principles engineering logic. - Transcend says that approach is designed to keep every calculation tied to established engineering principles and a utility’s own design standards. - TDG can generate preliminary designs for wastewater and drinking water treatment plants, electrical substations and pump stations. - Users enter basic project criteria such as flows, loads, treatment objectives and site constraints. - TDG produces complete preliminary designs in minutes. - The system runs process, civil, mechanical, electrical and structural calculations in one engine. - When one assumption changes, TDG updates all disciplines together. - Each design package can include process diagrams, 3D building information modeling, equipment lists, bills of quantities, operating expenditure estimates and a basis of design. - Ask TDG answers plain-language questions about engineering methods, calculations and technology-selection approaches used in TDG. - Ask TDG responses are grounded in TDG’s technical documentation. - Ask TDG does not change engineering results, perform calculations or replace expert review. - MCP Gateway allows a utility’s AI to call selected TDG capabilities, and TDG performs the engineering while the results flow into analytics, digital initiatives and downstream workflows. - Both capabilities sit on the same governed TDG platform. - TDG Nexus, Transcend’s ecosystem for sharing approved utility standards and technology specification packages, supports the platform foundation. - Utilities can share their own design standards packages through Nexus. - Technology specification packages from OEMs such as Xylem and DuPont can also be applied to designs. - Transcend says this creates a shared, traceable basis for evaluating options earlier in planning.

Between the lines: - The company is trying to make AI useful for engineering work without letting AI become the source of engineering truth. - That positioning matters for utilities, where early design decisions can influence budgets, scope and long-term asset performance. - The release also suggests Transcend wants to become the system of record for both engineering logic and AI-assisted workflows around that logic.

What’s next: - Utility, OEM and engineering teams can book consultations for Ask TDG at Ask TDG. - Enterprise AI, digital engineering and strategic partner teams can request an MCP Gateway briefing at MCP Gateway. - Transcend is likely to use the launch to deepen adoption among organizations that already run their own AI environments.

The bottom line: - Transcend is pairing deterministic engineering with governed enterprise AI so utilities can move faster without giving up traceability or expert control.