Company summary
PostAds Group is a privately-held marketing technology (MarTech) incubation company and parent company of Prodigy. It delivers business transformation using artificial intelligence (AI) and workflow innovations to enable global marketers to leverage the power of data for storytelling.
The company launched Prodigy with the SaaMS (Software as a Managed Service) workflow model to outperform legacy SaaS solutions by eliminating data entry burdens for subscribers.
The Challenge
Hindering growth with manual labor requirements for document scanning
Post Ads Group inputs thousands of customer marketing documents each month into large language models (LLMs) to train its AI and to generate high quality marketing output for its clients. It was processing over 10,000 customer documents per month, ranging from 3 to 15 pages per document. It was utilizing a template-based Optical Character Recognition (OCR) process that required a large amount of manual template design and data entry.
Each document took about 60 seconds to scan manually, a model that didn’t enable scale as Post Ads’ business grew. Additionally, the legacy template-based OCR system would often miscategorize content, requiring additional manual labor to create valid inputs. Post Ads needed a more efficient and cost-effective process for inputting over 70,000 pages of documents each month so it could deliver better data to its customers at scale.
The Solution
Creating an automated workflow to save time and human effort
To increase the automated process of documents and data in a faster and more cost-effective manner, Post Ads Group called on CloudHesive to design and implement an automated document processing solution in the AWS cloud.
The document processing solution enabled Post Ads Group to eliminate manual human intervention and speed up document processing time by using Amazon Textract, a machine learning (ML) service that automatically extracts text, handwriting, layout elements, and data from scanned documents.
CloudHesive built a workflow that used Amazon Simple Queue Service to queue, poll, and pull documents from Post Ads’ specified Ready folder, process and classify document text automatically and accurately using Amazon Textract, and deliver the final content to Review and Complete folders.
The benefits
Saving time and effort with machine learning automation in the AWS cloud
The document processing solution that CloudHesive built on the AWS cloud enabled Post Ads Group to entirely eliminate manual human intervention entirely and speed up document processing time.
The average scan time for each document dropped by 99%, from 60 seconds to just milliseconds using Amazon Textract.
Services that drive success
Devops
- Deployment of Amazon Textract, Amazon Simple Queue Service, and AWS Lambda for automated document scan workflow
- Built deployment scripts for environment deployment
Customer benefit summary
- Eliminated manual handling of documents entirely
- Decreased average document processing time by 99%, from 60 seconds to just milliseconds
AWS technologies used
Industry
Marketing and advertising
Segment
Private
Region
NAMER
