QuietBatch: let your to-do list move at a slower pace

sebasidian sunday — Mon, 26 Oct 2026 06:43:06 +0000

Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady
What I Built
QuietBatch groups a pasted task list for review. Paste the loose ends, one per line, and an open semantic embedding model suggests which tasks might belong together. It also surfaces possible repeats that use different words.

The result is a review, not an autonomous cleanup. You can accept a batch, keep its tasks separate, mark a pair as related, or keep both distinct. Every original stays in the exported review. QuietBatch never books, emails, deletes or executes anything.
Demo

Watch the QuietBatch demo on YouTube

The demo uses ten synthetic tasks and local MiniLM inference. It follows one continuous operation: paste tasks, run the real local model, inspect the suggested batches, accept one batch, and undo the decision. The interface also supports reviewing repeats and exporting every original; those actions are outside this short clip.

To reproduce it locally:
git clone https://github.com/sundaysebasidian-byte/quietbatch.git
cd quietbatch
npm ci --ignore-scripts
npm run setup:model
npm start

Open http://127.0.0.1:48317 and choose Try a sample.

The bundled sample produces three multi-task batches and one possible repeat. The two invoice lines have a cosine similarity of about 0.947. Two grocery-related lines remain separate at the chosen grouping threshold. That imperfect result is useful context: the model offers a starting point, and the person keeps the final judgment.
Code
Source code on GitHub

QuietBatch is a new standalone implementation created on October 20, 2026 UTC. Application code is MIT licensed, and the model and runtime licences are documented with the source. It does not reuse any of my earlier challenge projects.
How I Built It
The central component is all-MiniLM-L6-v2, using the Xenova ONNX conversion pinned to revision 751bff37182d3f1213fa05d7196b954e230abad9. Transformers.js runs the quantized model through ONNX Runtime on the local CPU.

Each line becomes a normalized, mean-pooled 384-dimensional vector. Complete-link clustering proposes groups when every cross-pair in a merged cluster meets a cosine threshold of 0.45. A separate threshold of 0.82 surfaces possible repeats. Those cutoffs are engineering defaults, not measured probabilities or an accuracy score.

An embedding model compares the meaning of short tasks. A representative original task becomes the batch title, keeping the interface grounded in what the person actually wrote.

The interface moves from input to a focused review workspace, with separate Batches and Possible repeats views. Returning to the input preserves decisions until the text changes.

The browser talks only to a loopback server. Model downloads are an explicit setup step, not a surprise during task processing. The app verifies local files, shows loading and per-task progress, and runs inference in a worker that can be terminated when cancelled. When files are missing, it stops with repair instructions instead of pretending a keyword fallback is AI.

The tests run the real model on repeated input, paraphrases, different meanings and long text. They also exercise missing-file recovery, cancellation, review, undo, export and mobile layouts. A synthetic dental paraphrase pair scored 0.589 while a dental-versus-plant pair scored 0.052 in the recorded run. These examples demonstrate the forward pass; they do not establish general accuracy.

The model sees at most 256 encoded token IDs per task. Long originals remain intact, but their later tokens may not influence a suggestion.

AI assistance was used for the code, tests, article, interface, and demo review.
Why Does Open Innovation Matter?
A to-do list can reveal ordinary but personal details: medical appointments, money owed, travel plans. Local inference means those lines can be compared without sending them to a hosted inference provider. After installing dependencies and downloading the public model, this build can run without an internet connection.

Open components also make the behavior inspectable. The model revision, tokenizer, vector dimensions, grouping logic and thresholds are all visible. Someone can reproduce a suggestion, inspect where it failed, or change the model and evaluate the consequences. There is no inference subscription or API key to maintain, though the user's own computer still uses resources.

For this week's theme, the ambition is deliberately small: make it easier to notice a few tasks worth doing together. I have not measured time savings with users, and I do not claim that a cosine score understands a person's priorities.

Pocket Field Trip: local AI with somewhere to stop

sebasidian sunday — Mon, 12 Oct 2026 06:59:46 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass.
What I Built
Pocket Field Trip helps you notice something on an ordinary walk. Instead of giving you another feed to scroll, it offers a small observation task, a few cues to take outside, and a place to record what you noticed when you return.
The idea is simple: the screen should be the shortest part of the experience. You do not need a special destination or an ambitious hike. A familiar street can become a field trip.
Demo

Try the working browser app:
https://sundaysebasidian-byte.github.io/
Choose a prompt, read the observation card, step away from the screen, and return to save a short note. The browser demo is the easiest way to try the project; an Android build is also in development.

Code
The source is public under the MIT license:
https://github.com/sundaysebasidian-byte/sundaysebasidian-byte.github.io
How I Built It
The project uses an open-weight text embedding model through Transformers.js and ONNX Runtime Web. Inference runs locally in the browser through WebAssembly. The model helps match the user's input to a bounded collection of observation cards, rather than relying on an unrestricted text generator.
This keeps the activities understandable and testable. The AI helps select an appropriate starting point; the person does the observing.
The published demo has been tested with real browser inference, producing 384-dimensional embeddings across a collection of 20 cards. The interface and note-saving flow have also undergone regression testing. AI coding assistants helped with implementation and review.
Why Does Open Innovation Matter?
A small outdoor tool should not require every interaction to travel to a hosted AI service. Local inference lets Pocket Field Trip keep the matching process on the user's device and avoids a paid inference API for each prompt.
Open components also make the behavior inspectable. Developers can examine the matching logic, replace the model, or adapt the observation cards. After the required assets are downloaded and available locally, inference does not depend on a remote model endpoint.
What I Learned
The most useful AI interaction here is deliberately short. More generated text does not necessarily make an outdoor experience better. A clear suggestion and a reliable place to save one observation can be enough.
The browser version is ready to try. Android testing is still based on an emulator, so I am not claiming broad physical-device validation. The next step is testing the experience on real walks and improving it from that feedback.

DEV Community: sebasidian sunday

QuietBatch: let your to-do list move at a slower pace

Open http://127.0.0.1:48317 and choose Try a sample.

Pocket Field Trip: local AI with somewhere to stop