Products / AI showcase
Agentic Automation

Operations agent

Multi-step agents that triage inbound email, update records across systems, and orchestrate approvals with clear audit trails.

Client
Logistics & 3PL provider
Industry
Logistics
Duration
14 weeks
12+ hrs
reclaimed per FTE per week
94%
of emails auto-triaged
31 min
avg approval, down from 4.2 hrs
01 · Problem

Where the team was stuck.

  • 01Operations coordinators triaged ~800 emails a day — quote requests, delivery updates, invoice questions — and manually copied data into three separate systems.
  • 02Simple approvals (rate exceptions, credit holds) waited hours because the right approver was on a truck or in a meeting.
  • 03Every mis-keyed reference number cost 20 minutes of investigation later.
02 · Approach

How we built it.

Stack
  • OpenAI function calling
  • Temporal
  • MCP
  • NetSuite + custom TMS
01

Classify, extract, act

Inbound emails are classified into 14 workflows. Structured extraction pulls PO numbers, dates, and rates directly into typed schemas — no free-text guessing.

02

Agents that can be overruled

Every action goes through a durable Temporal workflow. Coordinators see a queue of proposed actions and can approve, edit, or reject before anything hits NetSuite.

03

Approvals go where the approver is

Threshold-based approvals push to Slack, SMS, or email based on the approver's on-shift preference. One tap approves; the agent finishes the workflow.

03 · Outcomes

What changed.

The coordinator team stopped growing linearly with volume. In the six months after launch, shipment volume grew 22% while headcount stayed flat. The team now spends its time on exceptions and customer relationships instead of data entry.

  • 12+ hrs
    reclaimed per FTE per week
  • 94%
    of emails auto-triaged
  • 31 min
    avg approval, down from 4.2 hrs
04 · Demo preview

Agent trace

A rate-exception request from a customer email, resolved end-to-end.

Book a live demo
session preview · illustrative
User
Email in: "Need a rate on 2 pallets ATL → MIA for Thu pickup, standard dry van, PO #77341."
· Classified: quote_request · Extracted: origin=ATL, destination=MIA, pickup=2026-07-09, equipment=dry_van, po=77341
AI
Looked up customer contract — rate table applies. Base rate $412, fuel surcharge $58. Requires manager approval (below floor by 4%).
· Approval sent to R. Patel (on-shift) via Slack · approved in 6 min
AI
Quote emailed to customer, load tendered in TMS as #L-92418, calendar hold placed for dispatcher.

Want the same for your team?

Start a conversation