A visual design tool that simplifies data model creation for Amazon Timestream, a serverless time-series database. This project addressed a critical adoption barrier by helping developers—particularly those migrating from SQL databases—understand and correctly structure their data in a schemaless environment, reducing setup time from 90+ days to a matter of hours.
Goal: Reduce cognitive load for new customers, help SQL database users understand schemaless time-series data structure, increase completion rates for Batch Load and Scheduled Queries workflows, and accelerate time-to-value.
New customers struggled to adopt Amazon Timestream due to fundamental misunderstandings about schemaless time-series data structure. Developers with SQL backgrounds attempted to structure Timestream databases using traditional relational patterns, resulting in inefficient queries, failed implementations, and a setup process that took over 90 days.
Developers with SQL backgrounds instinctively tried to structure data relationally, leading to inefficient schema designs that would only reveal problems after data ingestion.
Many customers would realize their data model was incorrect only after ingesting data and attempting queries, forcing them to restart the entire process or abandon the service altogether.
Data model formats and templates that could help customers were buried deep in documentation, making it difficult for new users to find validated starting points.
Customers struggled to complete key workflows like Batch Load (data import) and Scheduled Queries (retention driver), directly impacting service adoption and customer retention.
Why it mattered: This friction was significantly impacting service adoption and customer retention. The 90+ day setup time wasn't due to technical complexity alone—it was cognitive overload from trying to apply the wrong mental model to a fundamentally different data structure.
The approach centered on two key principles: matching the customer's existing mental model (SQL databases) while using the interface itself as a teaching tool for time-series concepts. Rather than forcing customers to abandon their existing knowledge, I designed a bridge to new concepts.
I investigated the root cause of low adoption rates through support ticket analysis, regular meetings with Solution Architects who worked directly with customers, and customer journey mapping of the typical 90+ day setup process.
I chose a visual builder approach over alternatives (improved documentation, tutorials, CLI tools) because it could simultaneously guide and educate users. Using Figma and the AWS Console Design System library, I created prototypes that made abstract concepts concrete.
I collaborated closely with service engineers to understand backend constraints and data model validation requirements, console engineers to implement the visual interface, and solution architects to validate that the tool addressed real customer needs and aligned with best practices.
Internal testing with solution architects revealed that the visual relationship mapping needed to scale effectively for large schemas and that template categorization needed refinement based on common use cases.
The visual builder transforms abstract time-series concepts into concrete, understandable structures while preventing common errors. Rather than relying on external documentation, the tool embeds contextual learning directly into the workflow.
Pre-configured templates based on common use cases provide validated starting points that follow best practices.
Upload sample data and the system automatically analyzes structure and populates correct fields.
Interactive diagram showing how dimensions, measures, and time columns relate in a schemaless structure.
Pre-configured templates based on common time-series use cases and supported data model formats. These templates, previously hidden deep in documentation, provide developers with validated starting points that follow Timestream best practices. Templates are organized by use case (IoT, DevOps monitoring, application metrics, etc.) with contextual guidance about when to use each approach.
Demo available to view on desktop.
Customers can upload sample data or records, and the system automatically analyzes the structure and populates the correct data model fields. This removes guesswork and ensures proper dimensional vs. measure field categorization. The auto-detection handles various data formats and provides confidence scores for its suggestions, allowing customers to validate the analysis.
Demo available to view on desktop.
An interactive diagram that shows how dimensions (attributes), measures (metrics), and time columns relate within the schemaless structure. This visualization bridges the mental model gap for SQL-background developers by making abstract time-series concepts tangible and understandable. For complex schemas, progressive disclosure patterns and grouping mechanisms prevent clutter. Real-time validation warnings alert users when their structure might lead to query inefficiencies.
Demo available to view on desktop.
Design Impact: The visual builder doesn't just enable configuration—it teaches time-series best practices. By embedding education directly into the workflow, customers learn the right patterns while completing real tasks.
The visual data model builder transformed the Timestream onboarding experience, dramatically reducing setup time and improving completion rates for critical workflows.
Bottom Line: This project demonstrates how thoughtful UX design can address not just usability issues, but fundamental adoption barriers rooted in knowledge gaps and misconceptions. By focusing on the customer's mental model, we created a bridge to new technology rather than expecting customers to abandon their existing frameworks.
Breaking down complex, unfamiliar concepts into visual, guided experiences dramatically lowers barriers to entry for new users. The 90+ day setup time wasn't due to technical complexity alone—it was cognitive overload. Visual representation made abstract concepts concrete while preventing common structural errors.
Understanding a customer's existing mental model (SQL databases) and their intent (data migration) allowed me to design an experience that bridged the gap to new concepts (schemaless time-series) rather than forcing them to start from scratch. Used familiar SQL terminology where appropriate but provided clear comparisons showing "SQL approach vs. Timestream approach."
Rather than relying on external documentation or tutorials, embedding contextual learning directly into the workflow ensured customers learned the right patterns while completing real tasks. The visual builder didn't just enable configuration—it taught time-series best practices through real-time validation and contextual guidance.
Close collaboration with Solution Architects who worked directly with new customers during onboarding provided invaluable insight into real-world pain points. Their input ensured the solution addressed actual customer needs rather than perceived problems, and their validation during testing confirmed the approach would work in practice.
This project reinforced the importance of understanding not just what customers are trying to do, but how they think about the problem. By designing an interface that met customers at their existing level of understanding (SQL) and guided them to new concepts (time-series), we eliminated a critical adoption barrier. Future improvements could include: expanded template library based on emerging use cases, collaborative data model design for teams, and version control for data model iterations.