Confidential client, name withheld under NDA

AI-Powered Real-Time Shipment Tracking for a 3PL

Quick answer: AI-powered real-time shipment tracking merges carrier, EDI and electronic logging device (ELD) data into one live view for delivery visibility, predicts ETAs with machine learning and flags late loads before they miss delivery appointments. Bacancy Technology delivered this for a US third-party logistics (3PL) provider.

Outcomes deliveredDelivered February to August 2026 by a team of 6
96%

More loads tracked in real time

One view across 350+ carriers

91%

More FTL ETAs within 30 minutes

ETAs refreshed on every ping

72%

Fewer manual check calls to carriers

Down from 1,200+ calls a day

4+ hrs

Advance warning on late loads

Fewer retail OTIF chargebacks

Client name withheld under NDA. To honour our non-disclosure agreement and compliance obligations, we describe this client by industry and region only. The project scope, technology and outcomes are shared without details that could identify them.

About the Client

The client is a Midwest-based US third-party logistics (3PL) provider that moves full truckload (FTL) and less-than-truckload (LTL) freight for retail and consumer goods shippers. It manages about 14,000 loads a month with more than 350 carriers.

The team tracked loads through check calls, emails and carrier portals, and ETAs were set at tender and rarely updated after pickup. Late deliveries surfaced only when receivers called. Any new system also had to work with the client's existing transportation management system (TMS), and small carriers could not be asked to install new software.

A 6-member Bacancy Technology team built an AI-powered real-time shipment tracking system between February 2026 and August 2026. Apache Kafka collects every tracking signal, XGBoost models predict ETAs and an AI exception agent flags at-risk loads. Live tracking now covers 96% of loads, and manual check calls fell by 72%.

Client type
3PL provider
Industry
Logistics
Location
USA
Team
6 resources
Timeline
February 2026 to August 2026
Project size
Non-disclosable
Engagement
Custom AI development

Four problems slowing delivery visibility

Scattered carrier data and fixed ETAs kept the team chasing loads instead of managing them.

Scattered carrier updates

What was happening

More than 350 carriers sent status updates through EDI 214 messages, APIs, ELD pings and emails.

Business Impact: Only 58% of loads had usable tracking at any time.

Static ETAs

What was happening

ETAs were set at tender and rarely updated after pickup.

Business Impact: Shippers and receivers planned around arrival times that were often wrong.

Delays found too late

What was happening

Delays surfaced only when receivers called about a missed appointment.

Business Impact: Missed dock slots led to detention disputes and on-time in-full (OTIF) chargebacks.

No self-serve visibility

What was happening

Shippers had no tracking portal, and staff had no single view of active loads.

Business Impact: Staff made more than 1,200 check calls a day to answer status questions.

Challenges & Solutions

Our data and AI engineers built the tracking system in phases around the client's existing TMS.

Challenge 1: Bringing 350+ carrier data feeds into one view

Every carrier sent updates differently, through EDI 214 messages, APIs, ELD pings or emails.

Solution: As a Logistics Application Development Company, we built an Apache Kafka data layer that collects every tracking signal. We also added an SMS link for drivers without an ELD and matched each update to its load by BOL and PRO number.

Challenge 2: Replacing static ETAs with predicted ones

ETAs ignored traffic, weather, drivers' remaining hours of service and dwell time at docks.

Solution: Our data scientists trained XGBoost models on three years of the client's lane history. Each ETA is recalculated on every ping and comes with a confidence range.

Challenge 3: Flagging late loads before the receiver calls

The team had no early signal to rebook dock appointments or warn shippers.

Solution: Our AI agent developers built an exception agent on LangGraph, powered by Claude on Amazon Bedrock, that flags at-risk loads. It drafts carrier check-ins and shipper delay notices, and staff approve every message before it is sent.

Challenge 4: Giving shippers and staff one live view

Shippers wanted self-serve tracking, and the system had to work with the existing TMS.

Solution: Our React developers built a React control tower and a shipper portal with Mapbox live maps and tracking links. Updated ETAs flow back to the TMS through its API. Shipper status calls and emails fell 45%.

How the Platform Works

AI-Powered Real-Time Shipment Tracking for a 3PL

Core Features

  • Multi-source shipment trackingEDI, carrier API, ELD and SMS data in one timeline per load.
  • Predictive ETA engineETAs with confidence ranges, updated on every ping.
  • AI exception agentAt-risk loads flagged with drafted carrier and shipper messages.
  • Operations control towerLive map of every active load with exceptions first.
  • Shipper tracking portalSelf-serve tracking links and live ETAs for every shipment.

The stack behind the tracking system

Backend
Python | FastAPI
Frontend
React | Mapbox GL JS
Data streaming
Apache Kafka | Amazon MSK
ML and ETA prediction
XGBoost | Amazon SageMaker
AI agent
LangGraph | Claude on Amazon Bedrock
Database
PostgreSQL | PostGIS | Redis
Integrations
EDI 214 | Carrier REST APIs | ELD APIs | Twilio SMS | TMS API
Cloud and DevOps
Amazon EKS | Docker | Terraform

Ready to start?

2500+ projects delivered by Bacancy Technology

Get access to an experienced team of developers and engineers from Bacancy Technology, handpicked to ace your goals. Kickstart within 48 hours, no-risk trial.

Book A 30 Min Call
14+Years of business experience
1458+Happy customers
30+Countries with happy customers
1050+Agile-enabled employees

More Case Studies

View All Case Studies