SHIPSY / AGENTFLEET / INITIALIZE LIVE

Gartner Symposium · Barcelona 2026 · Booth #521

Logistics, in its
System of Action era.

The AI-native execution layer for the autonomous supply chain.

See AgentFleet Mission Control diagnostic →

scroll

The category we're building

System of Record→ System of Action.

Every enterprise has a TMS, WMS, ERP. None of them act. Shipsy is the layer that does.

System of Record

Tells you what happened.

Logs shipments after the fact
Surfaces exceptions on a dashboard
Generates reports nobody reads on Monday
Waits for a human to do something

passive · descriptive · post-hoc

System of Action

Resolves it before you see it.

Reads context across every system you run
Detects the exception in <90 seconds
Acts — reroutes, reconciles, replies, replans
Only escalates the edges that need judgement

live · prescriptive · executing now

Monday morning, on Shipsy

Outcomes you can put a date on.

0% Avg cost reduction Across transportation spend

$0M Largest annual saving Single customer, annualised

0% WISMO reduction Via proactive AI notifications

0% Driver productivity lift Through AI route optimisation

From plant gate to store shelf

AI that executes the complete CPG supply chain.

CPG supply chains span dozens of suppliers, hundreds of distribution points, and millions of deliveries. Shipsy replaces fragmented, manual execution with a single AI-native execution layer — from inbound milk runs and FTL optimisation to last-mile store replenishment and freight settlement.

50%fewer failed deliveries

28%LSP cost reduction

$5M+single-deployment saving

Multi-modal planning & FTL optimisation Driver-first execution & e-POD Real-time control tower Returns management Automated freight settlement

Customers HEINEKEN Coca-Cola HCCB Jain Irrigation Kellogg's

Carrier → depot → doorstep

Courier and logistics networks that run on AI, not spreadsheets.

Shipsy reimagines CEP operations with AI-powered orchestration that optimises every step — depot sort, intelligent last-mile routing, proactive customer notifications, multi-carrier orchestration — at the scale of billions of annual shipments.

20%+driver productivity lift

60%depot time reduction (120 → 50 min)

5 Bn+shipments processed / year

Guided sort & loading Intelligent last-mile routing Proactive CX automation Multi-carrier orchestration Real-time network visibility

Customers Aramex Teleport Maldives Post Qatar Post Nippon Express

Warehouse to doorstep, every promise kept

Retail logistics that delivers on every promise.

From omnichannel distribution to D2C last-mile, Shipsy gives retail and e-commerce brands the AI infrastructure to cut cost-per-shipment, optimise across carriers, and deliver an exceptional post-purchase experience — across urban metros and tier-2/3 markets alike.

91%on-time delivery rate (Wellness Forever)

41%cost improvement

59 minavg delivery time

AI route optimisation Smart carrier allocation Post-purchase CX & tracking Returns automation Multi-channel visibility

Customers Decathlon GMG Wellness Forever Yum! Brands Spencer's

Cold chain to patient · compliance-first

Pharma logistics built on auditable AI.

Healthcare SLAs, cold-chain compliance, and audit-grade chain of custody — orchestrated on top of whatever Oracle, SAP, or custom SOR you already run. Settlements, disputes, and exception comms automated with three-tier safety calibrated for healthcare and customs.

Oraclenative integration · Arginx live

99%+cold-chain SLA adherence (Catalent)

30+countries — Arginx Pan-region

Cold-chain & hazmat compliance Oracle / SAP-native settlements Audit-grade chain of custody Dispute auto-reconciliation Three-tier safety (healthcare-tuned)

The Oracle proof point

Arginx — global biopharma, $45B+ market cap, 30+ countries — runs freight settlements end-to-end on AgentFleet via native Oracle integration. Oracle SOR untouched; Shipsy reads, reconciles, and writes back decisions. Zero compliance incidents in 18 months of autonomous cold-chain operation.

Customers Arginx Catalent Apollo Pharmacy Wellness Forever

AgentFleet

Five agents. Live on enterprise customers.

Click any agent on the left to see how they spend their day.

Auto-cycling — click any badge to pause

Autonomous Control Tower

Purpose-built agent · Atlas

Day in the life

Replans the dispatch board continuously as ETAs drift, capacity changes, and exceptions land.
Reroutes around closures, weather, port congestion — issues drivers the new plan before they hit traffic.
Evaluates FTL bids across all carriers on a lane and books the optimal allocation under contract.

Plugs into

Your TMS Carrier APIs Mapping & traffic Customer SLAs Contract pricing

65% faster problem resolution (HEINEKEN, HCCB)

Customer Experience

Purpose-built agent · Clara

Day in the life

Answers retailer queries (status, POD, invoice attachments) end-to-end without human handoff.
Surfaces ETA shifts proactively to customers via WhatsApp / email / SMS before SLA breach.
Triages exception comms — escalates only what genuinely needs a human voice to the right team.

Plugs into

CRM Customer portals WhatsApp / SMS / Email Ticketing POD archives

30% reduction in inbound support volume (HCCB)

Freight Settlement

Purpose-built agent · Nexa

Day in the life

Parses freight invoices against contract clauses (rate, accessorial, fuel surcharge, weight tolerance).
Reconciles carrier payments line by line; pushes matched invoices straight to auto-pay.
Flags variances with the clause reference and holds disputes for human review.

Plugs into

Your ERP AP system Carrier contracts Invoice OCR Rate tables

40–50% reduction in manual settlement effort

Dispute Resolution

Purpose-built agent · Vera

Day in the life

Reconciles RPM disputes line-by-line against GPS, POD, and carrier rate-card records.
Triggers credit notes automatically when discrepancies are clear-cut.
Escalates only the genuine edge cases to a human reviewer with full context attached.

Plugs into

GPS / telematics POD archive Rate-card master Carrier portals Credit-note system

25% dispute cycle time reduction

Driver Assist & Dispatch

Purpose-built agent · Astra

Day in the life

Guides drivers through loading at the dock — sequence, weight check, paperwork — voice-led.
Issues real-time detours when closures hit, before dispatch even knows there's a problem.
Captures e-POD on every delivery and pushes it to whoever needs it within seconds.

Plugs into

Driver app Voice channel Mapping & traffic e-POD capture Dispatch board

15–20% driver productivity lift · 45–50% faster onboarding

Why Shipsy

More than AI. A deployment commitment.

The platform, the people who put it live, and the answers your IT lead will ask for — in one place.

The pillars

Built to Act, Not Record

The logistics industry runs on software designed to track shipments and generate reports. Shipsy is purpose-built as a system of action — every workflow is designed for autonomous execution, not just visibility.

Systems that record → systems that act

AI-Native Architecture

Most platforms bolted AI onto legacy infrastructure. Shipsy was built AI-native from day one — intelligence is at the core of every workflow, so AI doesn't just surface insights, it closes loops.

250+ enterprise customers · 30+ countries

AgentFleet: Your AI Workforce

Five specialised AI agents — Atlas, Clara, Nexa, Vera, Astra — each with defined roles, SOPs, and bounded autonomy. Humans govern the exceptions that need judgement; everything else runs automatically.

Atlas · Clara · Nexa · Vera · Astra

How we put it live · the FDE model

Field Deployment Engineers on-site with your team. Not remote support tickets.

Every Shipsy deployment includes a dedicated FDE pod — engineers embedded with your ops floor for the first 8–12 weeks. Fast iteration loops, real feedback, no time-zone gap between “something is off” and “it’s fixed in production”.

01

Weeks 1–2 · Discovery & mapping

Workflow mapping. Highest-ROI beachhead identified. Tribal-knowledge extraction starts. No platform commitment yet.
02

Weeks 2–4 · Integration & training

SOR connectors deployed. Ontology calibrated to your data model. Eval sets built and gated before any agent acts.
03

Weeks 4–8 · Controlled pilot

One workflow, limited scope. 10–20% agent autonomy, high HITL. KPI baseline established before any expansion.
04

Month 3+ · Expand & compound

Autonomy to 50–80% on the proven workflow. More departments onboarded. ROI gate must be passed before each expansion.

For IT & CISO · the six questions, answered

How much data?

Use-case scoped. Read-only by default; write access agent-by-agent.

RBAC mirrors humans

How private?

No training on your inputs. PII stripped pre-LLM in 90% of flows. EU / APAC / US residency.

Contract-enforced

Safety net?

Pre-deploy evals · live drift detection · circuit breakers · full audit trail.

Zero incidents · 18 mo

Our TMS/WMS/ERP?

Untouched. 20+ pre-built connectors. SAP-native, Oracle, Manhattan, custom.

No migration

Time to live?

4–8 weeks to first production agent. One use case. Then expand.

Dept by dept

Humans-in-the-loop?

Mobile-first push with one-tap approval. 15-min escalation. Routine never reaches you.

Edges only

Mission Control Diagnostic

See your Control Tower.

Three answers. One tailored tour.

Step 01 / 03 Your industry

What kind of operation do you run?

Step 02 / 03 Your pain

Where does the pain hit hardest?

Step 03 / 03 Your 12-month target

What does success look like in 12 months?

SHIPSY / AGENTFLEET / INITIALIZE LIVE

Your Control Tower · tailored to your operation live · agentfleet active

Your tailored demo is ready

Loading your view…

Opening in a new tab.

Talk to us at Booth #521 →

Analyst recognition

Three years on the Gartner® Magic Quadrant™ for TMS.

2024 · 2025 · 2026 · our customers have been running on agentic AI for three years.

2019

First Gartner mention

AI-driven last-mile delivery

2021

Gartner Cool Vendor

Multi-modal orchestration

2024

MQ for TMS · Year 1

First inclusion

2025

MQ for TMS · Year 2

Expanded EMEA / APAC

2026

MQ for TMS · Year 3

Live at Symposium Barcelona

Logistics has long run on systems that show you what went wrong after the fact. We're building the layer that closes the loop — where AI detects the exception, coordinates across carriers, drivers, and customers, and resolves it.

Soham Chokshi

Co-founder & CEO, Shipsy

Under the hood

Sits on top of your stack. Acts across it.

Your TMS, WMS, ERP keep recording. Shipsy is the layer that decides and executes.

SYSTEM OF ACTION ARCHITECTURE

01

Agent Layer

AgentFleet — autonomous logistics operators

AtlasClaraNexaVeraAstra

02

Intelligence Layer

Multi-LLM routing · Memory engine · Reasoning

ClaudeGeminiGPT-4DeepSeek

03

Ontology Layer

Logistics knowledge graph · Master data resolver

OTM SchemaEntity DedupSLA Encoding

04

Connector Layer

Pre-built adapters · SAP native bypass · RBAC

SAPOracle TMSManhattan WMS20+ more

SOR · Your System of Record No migration. No rip-and-replace.

SAP ERP / TM via SAP-native bypass

Oracle TMS live at Arginx

Manhattan WMS pre-built adapter

Any legacy REST · SOAP · EDI · SFTP

The bridge, visualised

Your SOR keeps recording. Shipsy acts.

Click a System of Record below. Shipsy reads the events, reasons across them, and writes the decision back — idempotent, replayable, audited. No re-platforming. No frozen roadmap.

Explore the full architecture

Barcelona · May 2026