Build agents that work.
Built on the Pedestal logistics graph, quickly launch agents that work with your enterprise data.
Example agents
Pedestal provides off‑the‑shelf agents for operations.
Each one goes past the recommendation and finishes the work: it makes the call, gets the approval, updates the system, and brings in a person when something is unusual.
Customer coordination
Calls and texts every customer before delivery to confirm the window and collect access notes. Live today at a Fortune 500 ecommerce company.
Labor optimization
Plans tomorrow’s staffing, gets the shift manager’s approval, and writes the plan back to your WMS.
Predictive maintenance
Spots equipment that is about to fail, ranks the risk across the fleet, and escalates each one, from a work order to a call to the ship’s captain.
Order processing
Reads purchase orders from email, PDF, and EDI, checks every line against price and stock, and enters the order in your ERP. Anything unusual goes to a person.
Operations analyst
An always-on data scientist that watches your operation for defects, anomalies, and issues, finds the cause, and flags it before it spreads.
Build your own agent
Have work only your operation does? Pedestal builds the agent with your team, from first idea to production in about a month.
The Pedestal process
Pedestal will build the agents you need.
Start from one of these, or from work only your operation does. We build it with your team on the Pedestal logistics graph, from first idea to production in about a month.
- 1
1 hour
Pick the use case
One session with your team to choose the first work to hand an agent.
- 2
1 week
Proof of concept
A working agent built on your anonymized data, for your team to try.
1 month
Live in production
Running in your live workflows and environment, with its impact measured.
Customer coordination
5,000+ calls a day for a Fortune 500 ecommerce company.
With a strong data foundation in place, Pedestal built voice and text agents that support every delivery and engage millions of customers a year, helping bring missed deliveries down by 20%+. Read the customer story
- Order systemArden sofa, 2 boxes, 2‑person crew
- Route planThursday, 8–11 AM, stop 14 of 22
- Delivery historyMissed a delivery in March
- Address notesGated community, code needed
- Delivery playbookCan move up to 3 days later
- Pedestal agent0:00Hi Nora, I’m calling about your Arden sofa. It’s set for Thursday between 8 and 11 AM. Will someone be home?Drew on:Order systemRoute plan
- Nora0:07Thursday doesn’t work. We’re out all day.
- Pedestal agent0:10No problem. We missed you last time, so let’s pick a day that works. Friday or Saturday, 8 to 11?Drew on:Delivery historyDelivery playbook
- Nora0:17Friday is great.
- Pedestal agent0:19Done, Friday 8 to 11. Your community has a gate. Is there a code the driver should use?Drew on:Address notes
- Nora0:25It’s 4482.
- Pedestal agent0:27Got it. Two people will bring both boxes in, and you’ll get a text that morning.Drew on:Order system
Systems updated
3 of 3
- Moved to Friday, 8–11 AMRoute plan
- Gate code 4482 added to the stopDriver app
- Confirmation sent to NoraText message
- calls a day
- 5,000+
- of customers reached a year
- Millions
- fewer missed deliveries
- 20%+
Labor optimization
Tomorrow’s staffing, planned, approved, and in your WMS before the shift.
The agent compares tomorrow’s work with who is scheduled and recommends the moves that close the gap. Then it keeps going: it asks the shift manager to approve in Slack and writes the approved plan back to your WMS.
- 1Recommends
Tuesday, first shift
Columbus warehouse
- 2Gets approval
#columbus-shift-leads
Slack
Labor agent5:30 PM
@Dana Ortiz 3 staffing changes for tomorrow’s first shift. Outbound is 22% above plan.
Press it as Dana to see step 3.Dana Ortiz approved all 3 5:42 PM
- 3Writes it back
WMS labor plan
Waiting for approval
- Outbound pickers, 6–10 AMWMS labor plan18→22
- Packers on overtime, 2–4 PMWMS labor plan0→2
- Dock 3 unloading startsDock schedule7 AM→1 PM
- Crew told their new startShift schedule6 people
Predictive maintenance
Catches equipment before it fails, and calls the captain when it can’t wait.
The agent watches sensor readings and maintenance records across the fleet, predicts what will fail and when, and escalates to match: a work order for routine fixes, a Slack message to the terminal team, and a phone call to the ship’s captain when cargo is at risk.
Build your own
Every agent is built on the Pedestal logistics graph.
It connects your systems with the rules your team works by, so each new agent starts where the last one did. We launch your first one together, in weeks. Build the next with us or on your own.
Agents
Pick one to see what it uses
- Your next agentYours to build
Logistics graph
Shared by every agent
- Orders
- Shipments
- People
- Equipment
Your rules
Confirm windows a day ahead
Managers approve staffing
Call the captain on cargo risk
Reads and writes back
Your systems
Nothing to replace
- WMS
- ERP
- TMS
- Maintenance
- Telematics
- Phone & text
- Slack & Teams
Every agent runs on Knowledge, the Context Layer, with the same Intelligence behind it.
Pick the work you’d hand over first.
Bring one repetitive process from your operation. We’ll show you the agent for it in the demo.