Example of Report
Example of Generating Chart
Example of generating Report

Snaply – Instant Reporting from Your Database

Inspiration

The idea for Snaply came from a familiar frustration: being just minutes away from a stakeholder meeting and realizing the necessary charts or reports aren’t ready. Creating them manually from the database usually takes too long. We wanted a solution that could remove this last-minute stress and deliver insights instantly.

What it does

Snaply transforms natural language requests into ready-to-share reports. It connects to multiple databases, generates SQL queries automatically, visualizes the results as charts, and compiles everything into polished PDF reports. Users can also schedule recurring reports by simply describing what they want and when. Reports can even be sent directly to email just tell Snaply “send it to your_email@gmail.com” and it will deliver the report instantly.

How we built it

Built a multi-DB SQL agent capable of connecting to different databases by using metadata.
Integrated an LLM to interpret user requests and generate schema-aware SQL queries.
Added visualization components to create charts and summaries directly from query results.
Developed a PDF report generator that compiles insights and visuals into shareable documents.
Extended the system with a scheduling agent, enabling hands-free, automated reporting.

Challenges we ran into

Ensuring SQL accuracy from LLM-generated queries required schema grounding and careful prompt design.
Embedding charts into PDF consistently was more challenging than expected.
Coordinating multiple agents (SQL agent, report agent, scheduling agent) under hackathon time constraints required strict prioritization.

Accomplishments that we're proud of

Delivering an end-to-end system that goes from natural language → SQL → charts → polished PDF reports.
Making Snaply intuitive enough that anyone can generate insights, not just technical users.
Building a scalable foundation for scheduling and automation within a short hackathon timeframe.

What we learned

The value of metadata-driven design in enabling LLMs to generate accurate SQL queries.
User experience matters: simplicity is key when dealing with complex backend systems.
Automation, like scheduling, transforms a reporting tool from reactive to proactive.

What's next for Snaply

Expanding chart types and report templates for richer customization.
Adding collaborative features, such as sharing reports directly with teams or Slack/Email integration.
Exploring real-time dashboards, so Snaply isn’t just for static reports but also for live insights.
Optimizing performance to handle larger datasets and enterprise-scale deployments.

All of databases using TiDB databases, include vector database. Email for TiDB Platform Account: affaalfiandy@gmail.com