LEN - turns sports photos into intelligent decisions with generative AI on AWS
Location
Latin America
Industry
Sports Technology / Generative AISolution
Personalized Insights Conversational AgentTechnology(ies)
Amazon Bedrock, AWS Lambda, Amazon SageMaker, Amazon API Gateway, Amazon S3
Company
LEN, a leading startup in sports photography content distribution in Latin America, set itself a new challenge: converting photographer data into useful and actionable knowledge.
Together with Amber, they developed a generative AI agent on AWS that interprets natural language questions and offers personalized insights on sales, performance, and audience behavior.
This innovation marks the beginning of the Photographer Intelligence-as-a-Service model, a pioneering approach that transforms transaction data into commercial intelligence for thousands of creators within the LEN ecosystem.
Challenge
LEN had valuable transactional information - sales, schedules, events, and tags - but lacked an automated and accessible way for each photographer to interpret their data and improve their performance.
The challenge was to create a tool capable of converting complex metrics into clear and personalized recommendations, without requiring technical knowledge.
Our proposal
Amber designed a serverless architecture on AWS with a conversational agent powered by generative AI, capable of analyzing performance data and offering personalized recommendations to each photographer.
The system transforms anonymized tabular information into practical and easy-to-understand insights, democratizing access to data intelligence within the LEN community.
Key technology components:
- Amazon Bedrock: Generation of responses and recommendations based on natural language.
- AWS Lambda: Serverless execution of queries and data processing.
- Amazon SageMaker: Training and deployment of personalized recommendation models.
- Amazon API Gateway: Secure and personalized access to the AI agent per user.
- Amazon S3: Centralized storage of data and analytical results.
Implementation process
The project was developed over 4 months using an agile methodology with 2-week sprints, prioritizing constant feedback from users.
Phase 1: Diagnosis
The need to transform scattered photographer data into useful and accessible knowledge, capable of generating individual value without manual analysis, was identified.
Phase 2: Design
A serverless architecture with generative AI on AWS was conceptualized, capable of analyzing metrics and responding in natural language with personalized recommendations.
Phase 3: Development
The conversational agent was implemented with Amazon Bedrock, SageMaker, and Lambda, integrating data in S3 and secure queries through API Gateway.
Phase 4: Validation
Real cases were tested with photographers, optimizing model accuracy and validating improvements in performance, satisfaction, and adoption within the LEN community.
Technical and business results
- Total automation in generating insights from transactional data.
- Real-time recommendations based on generative AI.
- Scalable serverless architecture with optimized costs.
- Rapid implementation validated by real users.
the LEN community
- Photographers empowered with information about what, how, and when they sell best.
- Greater retention and satisfaction of creators within the platform.
- Natural queries without prior technical knowledge.
- Consolidation of LEN as a benchmark for technological innovation in the region.
Testimonials
Tomás De Col CEO of LEN
"This tool allows us to help each photographer grow on our platform. They no longer just upload photos: now they can understand what works and what they can improve."
LEN partner photographer
"We went from having a simple sales dashboard to a conversation with an intelligent agent that tells you how to improve. That completely changes the experience."
Other success stories
We drive business efficiency with artificial intelligence agents that automate processes and deliver real results.
