back to main site
Echobot Journal is a personal project that combines my love for journaling with my interest in generative AI. It's an interface for journaling that uses OpenAI's API to generate insights and reflections from a user's journal entries. The project explores ways to process and interact with large volumes of personal data securely and locally, balancing privacy concerns with the benefits of AI-driven feedback.
The project originated from a desire to feed over 35,000 words of my own journal entries into AI while maintaining control over where that data was processed.
Journaling has been a lifelong practice for me—it's my mental health anchor, a way to reflect, refine my thoughts, and define my goals. With stacks of handwritten journals, I decided to digitise some of my entries and explore what AI could reveal.
However, I hesitated to share such personal insights directly with ChatGPT. This inspired me to build something local that could offer AI-driven reflections in a more controlled environment.
The Echobot Journal app operates in two parts: a Python backend and a Next.js frontend.
The backend uses FAISS for segmenting and indexing the journal text into manageable chunks. It serves as a locally run service to ensure more control over data processing.
The frontend is a clean and intuitive Next.js interface where users can upload journal entries, submit queries, and view AI-generated responses. Requests to the OpenAI API are made via the local backend, ensuring minimal data exposure.
Data is stored locally, and while it's not entirely private, it offers a more secure alternative to feeding data directly into a hosted service.
Next.js for the frontend interface.
FAISS and Python for text chunking and indexing.
OpenAI API for generative responses and insights.
Local storage for maintaining data control.
This project taught me how to work with the OpenAI API and process large volumes of text in a meaningful way.
It provided insights into building a locally hosted app while balancing user experience, performance, and privacy.
On a personal level, the generative AI responses highlighted patterns in my thinking, such as my ability to see the big picture but also becoming overwhelmed by its details—a valuable reflection that underscored the project's purpose of self-guided insight.
I'd love to continue developing Echobot Journal into a cloud-based app that others can use. While ChatGPT exists, I believe there's a unique opportunity to create a journaling-specific tool that offers guidance and structure for reflective practices.
Drawing on my lifelong journaling experience (I may have hit my 10,000 hours since my first pink journal with a lock and key at age 7!), I'd love to offer journaling prompts and techniques that align with AI insights to empower users in their reflective journeys.
The project is called 'Echobot Journal' because, at its core, it's about helping users hear their own thoughts and advice reflected back to them. It's a tool for uncovering patterns and wisdom already present, amplified through AI.