Data Whisperer – Giving Data a Human Voice

Inspiration

We’re surrounded by data, yet often disconnected from its human meaning. We wanted to build an app that gives data a voice – not just charts, but emotional narratives that foster empathy and awareness.

What it does

Data Whisperer transforms public datasets into short, AI-generated stories representing people behind the numbers. Users explore global issues like education or health through emotional micro-narratives.

How we built it

We used:

MongoDB Atlas for storing datasets and powering semantic + vector search.

Google Cloud (Vertex AI + Firebase) for story generation and app hosting.

GitLab CI/CD for code versioning, automation, and continuous deployment. One functional AI module was prepared for submission to the GitLab CI/CD Catalog.

Challenges we ran into

Mapping structured data into emotionally rich prompts.

Balancing factual accuracy with storytelling.

Ensuring privacy and avoiding synthetic bias in sensitive topics.

Learning and integrating multiple new platforms (MongoDB, Vertex AI, GitLab).

Accomplishments that we're proud of

Creating an MVP that makes global issues feel personal and urgent.

Building a system where data becomes storytelling.

Staying true to a deeply human concept in a tech-heavy challenge.

What we learned

How to combine AI with empathy.

How powerful MongoDB’s vector search is for narrative matching.

How GitLab’s DevSecOps flow improves team collaboration, even on small-scale creative projects.

What's next for Data Whisperer – Giving Data a Human Voice

Add multilingual support for localized emotional stories.

Let users upload their own data to “give it a voice.”

Expand into a digital empathy education platform for NGOs, journalists, and schools.

Integrate visual storytelling (images, audio) generated from the same data.

Built With

• chart.js
• express.js
• firebase
• gitlab-ci/cd
• google-cloud
• javascript
• mongodb-atlas
• node.js
• openai-api
• react
• tailwind-css
• vertex-ai