One epicenter for every agent we build.
A living home for the AISE agent portfolio — explore our featured agents, then dive into project showcases, training, and the Claude rollout.
Partner quote coming from the B2B team.
Built by Shelby Corbitt · Staff AI Solutions Engineer
Featured agents
Select an agent around the ring — its story pops up in the center.
IT Support Agent
The successor to TracE, AISE’s first production agent. Two agents, one for associates and one for technicians, that don’t just answer from vetted knowledge but act: tickets, device context, incident command, and live-agent escalation.
Open deep dive →IT Support Agent
The successor to TracE, AISE’s first production agent. Two agents, one for associates and one for technicians, that don’t just answer from vetted knowledge but act: tickets, device context, incident command, and live-agent escalation.
Open deep dive →Portfolio by stage
Every agent moves POC → Development → Test → Production under managed promotion.
AI/ML Initiatives
The Lean AI and Data Science pipeline beyond individual agents: forecasting, Project Helix, federal bids, and sustainability. These are our flagship initiatives, not a full list of everything the team is working on. Select a card for what it does and how we'll judge it.
PIPELINE AS OF 24 SEP 2026
How we build
Every agent follows the same deterministic, governed path from build to production.
Deterministic by design
Agents are grounded only in vetted Confluence and SharePoint sources, with structured tables preferred over unstructured PDFs.
Managed promotion
Builds move DEV → TEST → PROD as managed solutions, reviewed and approved by the AI Enablement lead — nothing reaches users unvetted.
Documented & governed
Every agent carries a governance document set, adapted to its platform — instruction sets, schema, security groups, logging, and testing/UAT.
Real teams, real deadlines, real outcomes.
Every project started as someone's actual problem. This is what happened when AISE pointed AI at it.
A $10M federal bid, made submittable
A USACE solicitation for Scott Air Force Base carried a ~$10M opportunity behind a compliance burden that normally makes bids like it too costly to chase. AISE assembled ~3,115 pages of GSA price-book evidence on a holiday deadline — putting MillerKnoll in the running.
From a $200K/yr buy decision to a bid engine in production
A VP was ready to license an external pricing tool. Instead, AISE and the Specials team built one together, and it went live on August 22. In its first 30 days it agreed with the quoters 91.7% of the time and helped cut quote response time from 72 hours to 26.1.
A 2,191-page admin guide, made instantly searchable
Parsed into 1,491 structured commands for exact lookup alongside semantic search — a retrieval layer built to validate itself with evals.
The doctrine that shaped every build since
Built and run by non-developers, processing thousands of knowledge articles — the hard lessons here became AISE’s deterministic, structured-knowledge-first doctrine.
Launch pad
Project Showcases
In-depth stories of flagship builds — AutoBid, IT Support, SAM, AI Intake, HR Studio, and the Data Assistant.
Open AutoBid showcase →Case Studies
Real MillerKnoll teams, real deadlines — in their own words where we have it.
Read the case studies →AI/ML Initiatives
The Lean AI and Data Science pipeline: forecasting, Project Helix, federal bids, retail and sustainability.
See the initiatives →Portfolio Impact
What’s in production, what it has saved, and who is using it.
See the impact →