This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend
What I Built
I built LifeLens, a privacy-first AI assistant designed for my mom and other family members who sometimes receive digital messages they don’t fully understand.
Bank notifications. Payment alerts. Delivery updates. Bills. Government messages. Important-looking links. Complicated English.
For me, they are easy to understand.
For someone who isn't deeply comfortable with digital technology, they can create a very simple question:
“What does this mean, and what am I supposed to do?”
Usually, the solution is to send me a screenshot.
That’s the problem I wanted to solve.
LifeLens lets someone upload a screenshot and get a simple explanation in the language they are comfortable with. Instead of showing a wall of technical text, it breaks the message down into:
- What is this?
- What does it mean in simple words?
- Do I need to do anything?
- Are there any warning signs I should know about?
The goal isn't to replace a bank, a support representative, or a security tool.
The goal is much smaller and much more human:
Make everyday digital information easier for the people we care about to understand.
How I Built It
The core idea behind LifeLens is local AI.
Instead of treating an AI API as a black box that receives someone's private screenshot, LifeLens is designed around locally running open/open-weight AI models.
The intended flow is:
Screenshot → Local vision model → Understanding → Simple explanation
For voice-based input, the same idea can be extended with local speech recognition:
Voice → Speech-to-text → Local AI → Simple explanation
The product is designed around a lightweight stack such as React, FastAPI, Ollama, local vision models, Whisper for speech recognition, and SQLite for local storage.
The important part isn't the framework choice.
It's where the AI runs.
A screenshot from a family member's phone can contain information about their finances, phone number, address, transactions, or other personal details. LifeLens is designed so that understanding that information doesn't have to mean automatically sending it to a third-party cloud AI service.
Why Does Open Innovation Matter?
This is where open innovation became more than a technical choice.
Imagine asking an AI to explain a bank message.
The message itself might contain private information.
With a closed cloud API, the natural architecture is:
Private message → Third-party server → AI → Response
With LifeLens, the goal is:
Private message → Your device → AI → Response
That difference matters.
Open/open-weight AI gives us the ability to experiment with models that can run locally, choose how inference happens, switch models, and build a product around privacy instead of treating privacy as an afterthought.
It also changes the economics of the product.
There doesn't need to be a paid API call for every screenshot.
There doesn't need to be an always-online dependency for the core experience.
And most importantly, there is a different kind of trust:
The AI can help you understand your information without automatically requiring your information to leave your device.
For me, that is what open innovation makes possible.
Not just another AI wrapper.
A different relationship between people, their data, and the AI helping them.
Top comments (0)