ParcelAgents

The parcel layer for your AI stack.

ParcelAgents are six purpose-built agents that audit invoices, file disputes, file claims, retrieve status, classify exceptions, and resolve inquiries across every major carrier. MCP-native. Token-priced. Drop them into Claude, your copilot, your ERP, or your own agents.

Get API access

>_ Read the docs

A dark themed code editor shows sample code for auditing invoices with parcelagents. Green text highlights: 1,284 line items audited, 47 overcharges flagged, $18,402 recoverable. Bullet points summarize disputes and median response time.

Some of the world’s best brands love ShipScience.

The image shows the word VIVO in bold black letters, with dots between each letter. The O is stylized as a swirling circular design with small dots inside.

Black text reads tires easy with a stylized tire tread graphic curving upward to the right of the text on a light background.

The image displays the word LUMENS in large, bold, black capital letters on a white background.

The image shows the word KENCO in bold, black, uppercase letters on a plain white background.

Every AI agent hits the same wall at the carrier.

Your team is building agents. Your vendors are building agents. Your customer's procurement team is building agents. They all work beautifully until one of them needs to do something real with a shipment: dispute a charge, file a claim, explain why a shipment is late, recover money UPS owes you.

Then it stops. Disputes live in portals. Claims need photos and receipts in carrier-specific formats. Status codes don't map. Inquiry workflows are phone trees with timeouts.

We've spent eight years on the other side of that wall. ParcelAgents is how you call through it.

Six agents. One API. Pay per task.

Shipment Fetch Agent

parcelagents.shipment_fetch

Pulls full charge-level detail from any carrier for one shipment or millions, by tracking number, account, date range, or any combination. Normalized to one query across all carriers.

Why this is hard

Every carrier reports charges differently and at different grains. UPS breaks out 30+ accessorial codes; FedEx rolls some into base; USPS uses a different taxonomy entirely.

Call it from: finance copilots · data warehouses · BI tools

Invoice Audit Agent

parcelagents.invoice_audit

Line-level audit of any carrier invoice. Catches overcharges, late deliveries, address corrections, dim weight errors, and surcharge anomalies.

Why this is hard

Surcharge logic changes annually and varies by zone, weight, dim factor, and accessorial stack. Our audit engine is the same one that's recovered millions for enterprise shippers.

Call it from: finance copilots · GL automation · dispute pipelines

Claims Filing Agent

parcelagents.file_claim

Files loss and damage claims with the right photos, receipts, and tracking documentation. Assembles documentation based on carrier requirements. Routes by carrier, brand, and declared value.

Why this is hard

Approval depends on documentation completeness and carrier-specific framing. Filing must occur within carrier filing windows and according to carrier submission requirements. Our agent files claims that get paid.

Call it from: CS automation · DTC support agents · 3PL claim flows

Claim Status Agent

parcelagents.claim_status

Retrieves and normalizes claim status across carriers. Translates carrier-specific states into one consistent schema. Provides details on claim lifecycle and the date the carrier last posted an update.

Why this is hard

"In review" means six different things across carriers. Carrier substatuses require translation, interpretation and normalization in order to provide value.

Call it from: Status pages · CS dashboards · ERP integrations

Dispute Filing Agent

parcelagents.file_dispute

Files and tracks billing disputes across UPS, FedEx, USPS, DHL, and regional carriers. Handles documentation, follow-up, and resubmission.

Why this is hard

Each carrier has a different dispute window, evidentiary standard, and rejection reason taxonomy. We've mapped them.

Call it from: Recovery automation · 3PL pass-through · Finance ops

Inquiry Management Agent

parcelagents.inquiry

Handles receiver, sender, and finance inquiries with carrier callback orchestration and structured resolution. Request invoice reviews or get clarification on questionable charges.

Why this is hard

Inquiry resolution often requires a phone call to a carrier rep. We handle the orchestration.

Call it from: Support deflection · Receiver self-service · Finance follow-up

Try out ParcelAgents with your AI tech stack → Get API Access

From key to recovery in four steps.

Provision a key

API key from your ShipScience support team.

Connect carriers

We handle credentials, rate limits, and carrier quirks.

Call from anywhere

Your app, copilot, Claude, ChatGPT, or your own agents.

Pay per task

Settle monthly or roll into your ShipScience contract.

Why this works when general-purpose AI doesn't.

ParcelPulse isn't a feed of generic alerts you'll never read. It's a structured cadence of PLD-based analysis, built around the moments that matter for your enterprise parcel program.

Parcel-native data underneath

A general agent with carrier API access still hallucinates. It doesn't know that a 0.04 dim divisor change costs you $40K a month, or that a FedEx Ground claim filed on day 61 is dead on arrival. ParcelAgents runs on the same shipment, surcharge, and claims schema that powers ActiveAudit and SimpleClaims — the platform that's processed over 100M shipments in the last 12 months.

Built for the AI stack you're actually using

MCP-compatible out of the box, with published tool-use schemas for Claude and OpenAI. SDKs in Python, Node, and Go. Or hit the REST API directly. Authentication, rate limits, retries, and carrier-specific quirks are handled for you, so your agents stay focused on the work instead of the plumbing.

Recovery rates you can put in a board deck

Enterprise customers see 3× recovery multiples on claims, 71% approval rates on large-shipper accounts, and median Year 1 recoveries of $216K on accounts shipping 50K+ parcels. ParcelAgents inherits that engine.

Deploy ParcelAgents using your data → Access the API

Designed for enterprise parcel shippers.

Companies that negotiate sophisticated contracts, ship millions of parcels annually, and have internal logistics teams — but still lack continuous intelligence about carrier changes.

Ecommerce Brands

Omnichannel retailers

3PLs / Warehouses

Industrials / Parts

Perishables / Subscriptions

Token-priced. Pay per task.

Per-call token costs are published in the developer docs. ShipScience platform customers get bundled token allowances at no extra cost.

Starter

ISV pilots, internal proofs

Blocks of 10,000 tokens
ISV Pilots, Internal Proof
REST + MCP access
All 6 agents
Community support

Enterprise

High-volume shippers, 3PLs, marketplaces

Everything in Starter
Custom token allowance
SDK + tool-use schemas
Priority Support + SLA
Bundled with ShipScience

Enterprise-grade security, built in.

Security and compliance features trusted by leading enterprises

SOC-2 Compliant

Role-based access

SSO (SAML/OIDC)

Encryption at rest / in transit

Audit Logs & SIEM export

Visit Trust Center

Answers to common questions.

Are ParcelAgents replacing other ShipScience products?

No. ParcelAgents is the engine underneath ActiveAudit, SimpleClaims, and ParcelPulse. The same agents we expose through the API are the ones that run the platform you may already be using. Customers who want a managed UI, dashboards, and a CSM use the platform.

Customers who want to embed parcel intelligence into their own product, build internal agents, or automate finance and CS workflows use the API. Most enterprise customers end up using both.

Can Claude or ChatGPT call these directly?

Yes. ParcelAgents is MCP-compatible out of the box, with published tool-use schemas for Claude and OpenAI. That means you can connect ParcelAgents to Claude Desktop, an internal Claude or GPT-powered copilot, or any agent framework that supports MCP or OpenAI tool-use.

Most customers start by giving their internal AI assistants the ability to call Shipment Fetch and Claim Status, then expand from there.

What carriers are supported?

UPS, FedEx, USPS, DHL Express, DHL eCommerce, OnTrac, GLS US, and Veho today. We're actively adding regional carriers based on customer demand. The agents return data in one normalized schema regardless of which carrier the call hits, so you don't write carrier-specific logic on your end. Full carrier list and per-agent capability matrix in the developer docs.

Do I need new carrier accounts?

No. ParcelAgents uses your existing carrier credentials. You connect each carrier once through your ShipScience admin, and we handle credential rotation, rate limits, retries, and carrier-specific quirks from there. No new contracts, no separate billing relationships, no rate renegotiation.

How does ParcelAgents compare to platform pricing?

The ShipScience platform is seat-based and bundled, sized for shippers who want a full audit, claims, and analytics workflow with a CSM behind it. ParcelAgents is pure usage. You pay per task, with token packs that scale from pilot to enterprise.

ISVs and 3PLs typically start on Growth, embed the agents into their product, and roll into a custom enterprise contract once volume justifies it. Platform customers get bundled token allowances at no extra cost, so you can experiment with API calls without changing your contract.

What does a typical token cost look like?

Per-call token costs vary by agent. Lightweight calls like Shipment Fetch and Claim Status are inexpensive and designed to be called at high volume. Heavier calls like Claim Filing and Dispute Filing cost more per call because they include documentation handling, carrier submission, and tracking.

Full per-agent token pricing is published in the developer docs so you can model spend before you commit.

How fast can we go live?

Most customers are calling the API within a day. Provisioning a key takes minutes. Connecting carrier accounts takes as long as it takes you to find the credentials.

The longest part is usually deciding which agent to call first. ISVs embedding ParcelAgents into a product typically ship a working integration in two to three weeks.

What about data security and compliance?

ParcelAgents runs on the same infrastructure as the ShipScience platform. SOC 2 Type II, encrypted in transit and at rest, with role-based access controls and full audit logs on every API call. Carrier credentials are stored in a dedicated secrets manager and never exposed to the calling application.

Start calling agents in under an hour.

Provision a key, connect your carriers, and ship your first integration the same day. Talk to sales if you'd rather scope it together.

Get API Access