How Intertec Automated Multilingual Voucher Generation with AI
Value Delivered
+100%
Reduction in error, ensuring consistent and accurate voucher titles, captions, and terms by strictly enforcing keyword and formatting rules
+100%
Increase in scalability, due to AWS allowing our solution to scale efficiently, vouchers and content can be generated on command, able to follow market demands
+80%
Decrease in production time of voucher generation
15+
Languages are currently being utilized in over 25 countries. It does not stop here as Intertec's solution supports over 100+ languages
INDUSTRY
E-commerce
PROJECT DURATION
10 years
LOCATION
Worldwide
CLUTCH REVIEW
5
Client Bio
An e-commerce platform needed to serve customers around the globe with accurate, localized content. But each department had its own system for translating product descriptions, support articles, and marketing materials. This patchwork approach led to inconsistencies and made it hard to maintain a single, unified brand message.
Situation
The client is recognized for offering a broad selection of discount vouchers across various industries such as fashion, travel, dining, and entertainment. Over time, partnerships with numerous brands and affiliates expanded the offerings and extended reach to new audiences globally. This growth, while positive, also introduced significant operational challenges.
The client’s editorial team spends countless hours creating voucher content manually. Each voucher demanded an eye-catching title, a compelling description, and clear terms and conditions. In addition, every piece of content had to include specific keywords to satisfy both brand and legal requirements. The involvement of multiple languages further complicated the process, as each language required unique rules and precise translations.
The manual approach began to slow down the operation, posing a risk of missed publishing deadlines and increased potential for human error. An automated solution became imperative to maintain or improve quality while scaling to meet the growing demand.
Solution
Solution
Intertec’s team delivered a system for automatic voucher generation using advanced language models trained on historical data.
- Data Management: Amazon S3 securely stored historical data, while AWS Athena enabled efficient querying and filtering for model training.
- Model Fine-tuning and Prompt Engineering: SageMaker’s GPUs fine-tuned pre-trained language models on the client’s dataset, with MLflow handling versioning and deployment.
- Deployment and Integration: Container images were stored in Amazon ECR and deployed via Amazon ECS with load-balanced APIs.
- Continuous Delivery and Monitoring: AWS CodePipeline and CodeBuild supported seamless updates, while CloudFormation, CloudWatch, and AutoScaling ensured rapid replication, real-time monitoring, and consistent performance.
Impact
- Multiple Languages, Countless Vouchers
Handling thousands of vouchers each month in various languages placed a heavy burden on the editorial team. Manual writing and translation led to delays and inconsistencies in the final output. - Keyword Rules and Formatting Requirements
Each voucher had to incorporate specific keywords and follow strict formatting guidelines that dictated capitalization, brand-specific language, and stylistic elements. Missing even one requirement could result in confusion or contractual issues with partners. - Growing Demand for Scalability
As the client expanded, a system capable of keeping pace without sacrificing quality was needed. Manual processes were unable to scale efficiently, creating a bottleneck in voucher creation and delaying the rollout of new deals.
