FreightLink is a Cognitive Automation Platform on the Cloud that bridges and automates the data and document exchanges between international trade stakeholders. Cognitive as in the decision engine driven by the latest Artificial Intelligence capabilities (NLP, OCR, ML, FIR) while Automation as in the execution of repetitive processes by computers (RPA). Combined on the cloud, it’s easily deployable, cost effective, intelligent, secured and take human out of any loop completely.
The key capabilities of the platform are:
1. Natural Language Processing (NLP): Aid computers to understand the human’s natural language and communications (to read, decipher and understand text).
2. Optical Character Recognition (OCR): Convert images of typed, handwritten or printed text into computer codes from any digital document.
3. Machine Learning (ML): Use scientific algorithms and statistical models to perform specific task without explicit instructions, relying on patterns and inference instead. It will replace human judgement.
4. Robotic Process Automation (RPA): Develop the action list by watching the user perform that task in the application’s Graphical User Interface (GUI), and then the computer will perform the automation by repeating those tasks directly in the GUI.
5. Facial/Image Recognition (FIR): Aid computers to understand faces and images (to recognise, classify and identify persons or objects).
Industry:
Global Freight Forwarder (GFF)
Background:
Before:
GFF ran a team of employees to search & download close to 8,000 Bill of Lading/Container (HBL/CNT) shipment data from 15 different partner websites across 10 countries daily and update these data into GFF freight management system (FMS) manually. Work and Information was frequently delayed or disrupted.
After:
Cognitive Automation Platform is implemented. The platform receives HBL/CNT data via RESTFUL API from GFF FMS. A few RPA bots will search & scrap data from the partner websites based on the HBL/CNT data. The RPA bots will update GFF FMS directly or via webservices depending on the completeness of data fetched (decided by machine learning). Manual data processing are taken out of the entire process. Greater stability, more speed and bigger throughput is achieved.
Project Duration:
8 weeks
Benefits:
a. The team of employees was redeployed
b. Accuracy increased significantly
c. Visibility on a single portal (Dashboard, Logged Activities & Exceptions)
d. Higher volume throughput to support business growth
e. Timely information is available
Industry:
Global Freight Forwarder (GFF)
Background:
Before:
GFF ran a team of employees to read close to 40,000 freight documents monthly (of approximately 900 different types/templates) and update these data into GFF FMS manually. Work and Information was frequently delayed or disrupted.
After:
Cognitive Automation Platform is implemented. Various digital Letter of Consignment, Import & Export documents, House Bill of Lading & Master Bill of Lading are uploaded into the platform via webservices. Machine Learning will recognise the various documents and work with OCR to extract the relevant data. These data are pushed into GFF FMS via RESTFUL API.
Various digital Booking Confirmation are sent to GFF email exchange. A few RPA bots will search the emails. NLP is used to recognise Booking Confirmation emails based on text recognition in the email subject and body. The digital Booking Confirmation are uploaded into the platform via webservices. Machine Learning will work with OCR to extract the relevant data. These data are updated into GFF via GFF FMS User Interface by RPA bots.
Manual data processing are taken out of the entire process. Greater stability, more speed and bigger throughput is achieved.
Project Duration:
4 months
Benefits:
a. Team of employees was redeployed
b. Accuracy increased significantly
c. Visibility on a single portal (Dashboard, Logged Activities & Exceptions)
d. Higher volume throughput to support business growth
e. Timely information is available
Industry:
Global Freight Forwarder (GFF)
Background:
Before:
GFF receive enquiries about their service routes, schedules, validities including prices via phone, emails or through web forms. It’s a tedious process to validate these customer requests, prepare sales quotations and respond to each request manually. And sales are lost if the quotations are not timely.
After:
Cognitive Automation Platform is implemented. All requests for quotations are raised via a chatbot, emails or web forms. A few RPA bots fetch customer requests from all the available channels and extract relevant data. NLP determines the intent of the requests and recommend actions dynamically. The RPA bots fetch the requested service information from GFF FMS through RESTFUL API, prepare sales quotations into pdf and respond to each request automatically. The platform also orchestrate sales quotation approval workflows before the RPA bot sends to customers. The platform provides a tailor-made dashboard.
Project Duration:
4 months
Benefits:
a. Accuracy increased significantly
b. Visibility on a single portal (Dashboards & Exceptions)
c. Higher volume throughput to support business growth
d. Timely customer responses
Industry:
Global Food Company (GFC)
Background:
Before:
GFC ran a team of employees to monitor the quality certifications & compliances including the operational and financial statuses of more than 1000 vendors globally through qualified web sources. This is critical to managing GFC supply chain risks. Information was frequently late and inaccurate.
After:
Cognitive Automation Platform is implemented. A few RPA bots fetch contents from qualified web sources 24/7 while Machine Learning classify them & recommend actions dynamically. Machine Learning learns through continuous training when users identify wrong classifications & recommended actions. Manual data processing are taken out of the entire process. Greater stability, more speed and bigger throughput is achieved.
Project Duration:
6 months
Benefits:
a. Team of employees was redeployed
b. Accuracy increased significantly
c. Visibility on a single portal (Dashboards & Exceptions)
d. Higher volume throughput to support business growth
e. Timely information is available