Case study
WokePeople
Generative AI on Amazon Bedrock that turns people data into analysis, with managed operations.
About the client
AI-powered executive search
Woke is a Brazilian company specializing in executive search, leadership development, organizational culture and career management. Founded in 2016, it combines technology, data science and innovation to empower people and companies going through transformation and the new economy.
Its services include:
- Executive search;
- Leadership development;
- Onboarding and integration programs;
- Organizational culture solutions.
Its goal is to transform working relationships through constant innovation and a focus on performance. Based in Brazil, Woke serves both established companies and fast-growing startups, and it maintains an active presence on Instagram, sharing content about its areas of expertise.
To make these solutions more efficient, KXC recommended artificial intelligence services from Amazon Web Services (AWS) that improve data analysis and the application's results.
The challenge
Data analysis at scale with artificial intelligence
WokePeople was looking for ways to improve its processes so that data analysis, combined with artificial intelligence services, would deliver insights more efficiently.
The large volume of data used in individual and group assessments made human analysis challenging, because it was hard to correlate so much information. The company needed tools that could analyze large volumes of data and turn them into information professionals could understand.
The demand called for a generative AI solution that kept the human touch that defines Woke's business model, with scalability and high availability.
The solution
Generative AI at scale, without losing the human touch
In the solution's main flow, Amazon API Gateway routes each request to one of two AWS Lambda functions, which invoke Amazon Bedrock to generate the analyses.
The second part of the project is an interactive chatbot built with Amazon Lex, an AWS Lambda function and integration with Amazon Bedrock, which makes it far more capable than a standard chatbot.
These services gave the infrastructure high availability and scalability, raising the quality of the application and the satisfaction of its users.
AWS services used
Results
A highly available, secure and scalable environment
The project delivered high availability, security, scalability and an overall improvement to the application. KXC met WokePeople's goals by supporting the company's growth, choosing the right AWS services and collaborating throughout the development of the solution.
The quality and speed of the analyses improved by 40% in how well WokePeople's clients understand them. With faster analyses, interactions with clients also move faster, which helps the company grow and stand out for quality.
Going forward, having more data stored in a database could improve these solutions even further.
Managed services contract
After implementing the generative AI solution, WokePeople hired KXC for the ongoing operational management of its entire AWS infrastructure, ensuring high availability and continuous improvement of the platform.
Scope of the managed services contract
24/7 monitoring and operations
KXC operates WokePeople's entire serverless infrastructure:
- Continuous monitoring of Amazon API Gateway, AWS Lambda, Amazon Bedrock and Amazon Lex
- Custom Amazon CloudWatch dashboards with AI metrics (inference latency, tokens consumed and success rate)
- Smart alerts for performance and cost anomalies
- Monitoring of integrations with clients' HR and CRM systems
Ongoing operational services
- Lambda function management: cold starts, memory optimization and timeout tuning
- Amazon Bedrock model updates as new versions are released
- Prompt management and Amazon Lex fine-tuning to improve the chatbot
- Backup and versioning of AI configurations through infrastructure as code (Terraform)
- Security management: IAM, VPC endpoints and encryption at rest and in transit
SLA-backed incident management
The contract sets service levels suited to a B2B executive search platform.
FinOps and AI optimization
- Monitoring of Amazon Bedrock inference costs
- Token and prompt optimization to improve cost efficiency
- Usage pattern analysis for Reserved Capacity
- Monthly reports on consumption and on the AI solution's return on investment (ROI)
Continuous platform evolution
- Implementation of new features requested by WokePeople
- Testing and deployment of new Amazon Bedrock foundation models
- Expanded integrations with clients' systems
- Training for the in-house team on AI best practices
Support and communication
- A dedicated technical support channel (Movidesk and WhatsApp)
- Biweekly check-in meetings
- Up-to-date technical documentation in the knowledge base
Managed services contract results
Since the managed services contract began:
- 99.95%
- AI platform availability
- 25%
- lower inference costs through prompt optimization
- Zero
- downtime during Bedrock model updates
- 2 hours
- mean time to resolve incidents
More case studies
Book a call with one of our specialists
Fill in your details and the service you need. We'll get in touch to find a time that works for you.