Ask Your Project a Question. Get an Answer.
Steel Nexus is the AI intelligence layer built across the entire Steel Shire ecosystem. Ask a natural language question about any corridor, crew, or status - and get an answer drawn directly from your live project data, not a report someone built last week.
The Intelligence Your Project Data Has Always Had.
Finally Accessible.
Your projects generate enormous amounts of data - daily reports, tract updates, inspection records, pipe inventory logs, change orders. Steel Nexus is the tool that makes all of it queryable in plain language, so the answers your team needs don’t require a report request or a data export.
Under the Hood.
Built for Enterprise Project Data.
Steel Nexus isn’t a chatbot layered on top of your data. It’s a purpose-built intelligence architecture - permission-aware, geographically enriched, and designed to return precise answers from thousands of project records without processing documents it has no business touching.
Permission-Based Access Architecture
Steel Nexus operates across a permission-based S3 bucket architecture - meaning it only surfaces data a user is cleared to see. A field inspector querying Steel Nexus gets answers drawn from their project scope. A contractor gets answers from their records only. An executive gets the full portfolio view. Access is enforced at the query level, not just the login screen - so there’s no risk of a contractor surfacing another contractor’s rates, a field user pulling executive financial data, or anyone accessing records outside their role. No document duplication required across buckets; permissions travel with the data.
OpenSearch Indexing & Top-Chunk Retrieval
Steel Nexus doesn’t process entire document sets every time a question is asked - that would be slow and expensive. Instead, every document ingested into the platform is indexed via OpenSearch and broken into retrievable chunks. When a question comes in, Steel Nexus retrieves only the most relevant chunks across the indexed corpus and surfaces a synthesized answer. The result is fast, cost-efficient, and precise - even across projects with thousands of documents, daily reports, and field records.
Enriched Metadata & Geographic Context Layer
Every document and data point ingested into Steel Nexus carries a metadata envelope - project ID, corridor, stationing range, date, data type, and source system. This means Steel Nexus doesn’t just find text that matches your question; it finds the right record in the right geographic context. Ask about a specific milepost range and Steel Nexus returns results scoped to that location - not every document that mentions the same pipe diameter. The geographic and project context layer is what separates Steel Nexus from a generic document search tool.
Connector-Based Data Ingestion - Your Documents Stay Where They Live
Steel Nexus ingests data through a structured connector architecture - not manual exports, not one-time imports, and not a migration project. Critically, your documents don’t need to move. Steel Nexus indexes across your existing S3 buckets and connected systems in place - no consolidation into a single repository, no restructuring how your files are stored, no conversation with IT about moving sensitive project data somewhere new. Each connector (Skyray, Cobra, Havoc, Arpium, and others) keeps the index current as data changes in the source system, so the intelligence layer is always working from live information. New connectors can be added as your stack grows - including custom-built connections to legacy systems, third-party platforms, and external document repositories your team already relies on. Your data infrastructure stays exactly as it is. Steel Nexus just makes it queryable.
Built for Every Stakeholder
Whether you’re trying to get a quick answer about a specific corridor or looking for early warning signals across a full portfolio, Steel Nexus is built around the kind of questions your team is actually asking.
One Intelligence Layer.
Every Steel Shire Product.
Steel Nexus is purpose-built to operate across the entire Steel Shire platform - Skyray, Arpium, Havoc, Cobra, Arsenal, and Blackbird. Every update, document, and data point stored anywhere in the ecosystem is searchable through Steel Nexus. Add a new product to your stack and it becomes queryable automatically - no setup, no data pipeline to configure.
Your Data, Wherever It Lives.
Steel Nexus is built to connect with the full Steel Shire ecosystem - but it’s also designed to extend beyond it. We work with internal IT teams to connect Steel Nexus to custom data sources, legacy project management systems, and external document repositories - so your team can query the full picture of your project data, not just what lives inside Steel Shire.
Results From the Field
A large midstream operator was drowning in competing spreadsheets across multiple ROW contractors. No one could agree on which data was current.