Skip to content
← Back to Our Work

Diversified Energy Company

In-tenant enterprise AI for a NYSE-listed energy operator

Diversified Energy Company architecture schematic

One example subsystem from this build. In most engagements we deliver the entire stack top to bottom — application code, Docker, infrastructure, deployment, and scaling — all in-house. Where a client already has platform teams in place, we work alongside them.

Project Summary

Enterprise AI rollout for Diversified Energy Company, a NYSE-listed energy operator, built entirely inside their own Microsoft Azure tenant. A Postgres and Snowflake MCP server, authenticated end to end with Microsoft Entra ID and default-deny access scopes, gives Copilot Studio and Azure AI Foundry assistants governed access to the company's production data. The first workflow is Production Accounting: desk analysts start the day with a scheduled digest of their own reports and ask ad-hoc questions in the Microsoft tools they already use.

The learning loop runs in-tenant. When an agent proposes a learning, the desk lead approves or rejects it before anything is stored, and approved knowledge is written to a licensed Knowledge Bank under the approver's own name, so the desk's expertise compounds instead of living in one person's head. Every call is audit-logged, and the engagement is structured as an umbrella the platform grows under, desk by desk.

Our role: Architecture & Engineering

What We Did

  • MCP server connecting Copilot Studio and Azure AI Foundry assistants to Postgres and Snowflake data
  • Runs entirely inside DEC's own Azure tenant on Container Apps — data never leaves their environment
  • Microsoft Entra ID authentication end to end with default-deny access scopes and a complete audit log
  • Scheduled morning desk digests plus ad-hoc questions for Production Accounting desks
  • In-tenant Knowledge Bank license: agent learnings are held as proposals until a desk lead approves them, then stored under the approver's name
  • Initial scope: Production Accounting desk views and statement automation, widening desk by desk

Want this in your own workflow?

Tell us which process you'd start with. We map it, install the loop, and prove it before it widens.

Plan Your Rollout