AI can search Oracle Aconex correspondence and transmittals when the mail, attachments and metadata are copied into a store your organisation controls, indexed for meaning, and queried with retrieval-augmented generation (RAG) that cites the original record. CDE Sync™ or CDE Backup™ brings that content into your own Azure tenancy, and AI Insights in the Utopiaverse™ answers questions across it, together with your other CDEs.
Why Aconex correspondence is so valuable
On many major projects, Aconex is the formal channel between client, contractor and supply chain. Requests for information, notices, instructions, submittals and transmittals all pass through it. That makes the correspondence register one of the most important records a project produces. It shows what was asked, what was answered, who knew what and when. It is also one of the hardest records to use, because it grows to tens of thousands of items and most searches rely on knowing the right keyword, mail type or reference number.
What makes it hard to search
Keyword search finds words. It does not find meaning. A question such as "what did the client say about changing the platform canopy design?" may be spread across a dozen mails using different words, with the key detail sitting in an attached PDF letter. Scanned attachments add another layer, because the text is locked in an image until OCR extracts it. And when a project uses Aconex for correspondence while design runs in ProjectWise or Autodesk Forma, the full answer usually crosses platforms.
How RAG works on correspondence
RAG combines search with a language model. The content is first broken into passages and indexed by meaning. When someone asks a question, the system retrieves the most relevant passages from the index and gives them to the model, which writes an answer grounded in that content and cites where each point came from. Our article on how RAG search works on construction documents explains the mechanics.
For correspondence, the metadata matters as much as the text. Mail type, sender, recipient, date, status and reference let the AI filter and order results, for example "notices from the client since March" or "transmittals issued to the electrical subcontractor". Mapping Aconex fields into a common schema keeps those filters reliable across projects.
How to set it up
Agree which Aconex projects, mail types and document registers are in scope, and who should be able to query them.
Choose the Azure tenancy and region, checking residency requirements against the global data region map.
Connect Aconex with a least-privilege account and replicate mail, attachments, documents and metadata into your tenancy.
Apply OCR to scanned attachments so their text is searchable.
Index the content and configure access in line with project permissions.
Pilot with real questions from commercial, design and delivery teams, checking every citation.
The Aconex to Azure Blob Storage guide covers the connection itself.
What teams ask once it is running
Commercial managers use it to trace the history of a variation or a delay event. Design managers find every RFI touching a particular interface. Project directors get a quick view of what the client has raised in the last month. Document controllers confirm which transmittal carried a given revision. Because each answer cites the original mail or document, people can open the source and check it, which is essential for anything contractual.
Build your own or use the Utopiaverse™
Once the content is in your tenancy, your team can build its own AI over it, for example with Azure AI Search and a model approved by your organisation, and some teams expose the index to AI assistants through an MCP server. That path suits organisations with an established data team. The alternative is AI Insights, which runs natively on Azure inside your tenancy and is ready to use across every connected CDE.
Aconex correspondence and the Utopiaverse™
Correspondence rarely tells the whole story on its own. In the Utopiaverse™, Aconex mail and documents sit beside drawings from ProjectWise or Forma, site records from Procore and working files in SharePoint, all in one connected record in your own Azure. AI Insights answers questions across every project and platform with cited sources, the file viewer opens any document from any CDE in one place, and dashboards show correspondence and document control activity across the portfolio.
That gives the organisation one view of what was said and issued on every project, and the record stays available after the Aconex project closes through CDE Archive™. Our article on enabling AI to read Oracle Aconex covers the wider Aconex record.
See AI search across Aconex in the monthly live demo, or talk to the team about your correspondence records.