DEV Community: E. Smith

Slothstep: a lazy-friendly app that turns "ugh, I'll do it later" into three tiny steps

E. Smith — Fri, 23 Oct 2026 14:43:22 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge: Week 3 – Slow and Steady.

What I Built

Slothstep is a tiny web app for the thing sloths do best: not panicking.

You type whatever you keep putting off ("clean the whole flat before Saturday", "write that difficult email to my boss", "sort the files on my laptop") and an open-weight model answers with exactly three steps:

one you can start right now and finish in two minutes or less,
one that follows naturally,
one that finishes the job.

Every step has a realistic time estimate, and you can tick them off as you go. The point is not a perfect plan. The point is that the first step is so small there is no excuse left not to start.

The effort it saves: the hardest part of a dreaded task is rarely the task, it is the staring. Slothstep replaces "where do I even begin?" with one concrete, two-minute action. (On my machine a plan takes about 25 seconds to generate with a 4B model.)

Demo

Live app (demo mode with sample plans, because a free host cannot run a model): https://slothstep.onrender.com/
Real thing: clone the repo, ollama pull gemma3:4b, pip install -r requirements.txt, python app.py
Code:
banana0000 / slothstep

Slothstep: turn a task you keep putting off into three tiny steps with an open-weight model (Gemma 3 + Ollama). Dash app.
🦥 Slothstep

Describe a task you keep putting off and get three tiny, realistic steps. The first one takes two minutes or less, so there is no excuse left not to start.

Built for the Hacktoberfest Open-Source AI Challenge (Week 3: Slow and Steady) with an open-weight model (Gemma 3) served by Ollama.

Live demo: https://slothstep.onrender.com

Features
- Three steps with time estimates, validated so the model cannot return malformed plans
- Tick steps off as you go
- Light and dark mode (remembered in the browser)
- A badge showing whether the model server is running
- Runs fully offline with a local model, or against any OpenAI-compatible host serving an open-weight model
Stack: Dash + Dash Mantine Components, Python.

Run locally
ollama pull gemma3:4b pip install -r requirements.txt python app.py
Open http://127.0.0.1:8050.

Configuration

Variable Default Purpose

OLLAMA_MODEL gemma3:4b Model name (Ollama tag, or model id on

…
View on GitHub

Variable	Default	Purpose
`OLLAMA_MODEL`	`gemma3:4b`	Model name (Ollama tag, or model id on

slothstep.onrender.com

The free host may take about a minute to wake up on the first visit.

How I Built It

UI: Dash with Dash Mantine Components (light and dark mode), a single Python file.
Model: Gemma 3 (4B) running locally through Ollama. The same code can also talk to any OpenAI-compatible host serving an open-weight model. The free hosted demo runs in a clearly labelled demo mode with sample plans, since a free host cannot run a model.
Reliable output: the model is asked for JSON only, then the app validates it (exactly three steps, non-empty text, positive integer minutes) and shows a friendly error instead of garbage if the model misbehaves.
Prompting: low temperature, a hard limit of two minutes for step one, and an explicit "do not invent missing details" instruction so it doesn't hallucinate a kitchen you don't have.
Deploy: a render.yaml Blueprint for a free Render web service running with DEMO_MODE=true. A real model needs a private Ollama service on a paid instance, which the repo's ollama/Dockerfile supports.

Why Open Innovation Matters Here

What people procrastinate on is personal: unanswered emails, messy folders, awkward conversations. I don't want those sent to a closed API.

Private by default: run it locally and your to-dos never leave your machine.
Works offline: no internet, no problem. Ollama and a 4B model are enough.
Set up once, free forever: no subscription, no per-token bill for something you use ten times a week.
Swappable: because Slothstep speaks both Ollama and the OpenAI-compatible API, moving from a local Gemma to a hosted open-weight model is two environment variables, not a rewrite.

A closed model might write slightly nicer sentences. But for a three-line plan, a small open model is more than good enough, and I keep control of my data.

Prize Categories

Best Use of Gemma: Gemma 3 is the model at the core of the app.
Best Use of Render: the app is deployed from a render.yaml Blueprint.

Thanks for reading, and please take it slow. 🦥

Receipt Box: a private receipt sorter for freelancers, powered by Gemma

E. Smith — Sat, 17 Oct 2026 07:43:55 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle

What I Built

Receipt Box is a receipt and invoice sorter for freelancers that runs entirely on your own laptop.

Every freelancer knows the end-of-quarter ritual: a folder of crumpled photos, PDF invoices in six currencies, and an accountant asking "which of these have VAT?" Receipt Box turns that folder into a clean expense dashboard in a few minutes. Drop in photos or PDFs and a vision model reads each one: vendor, date, total, VAT, currency, and a category such as Software, Travel or Meals. You get:

A dashboard: total spend, VAT to reclaim, spend by category and by month.
A VAT-by-quarter table, because that is the number a freelancer actually has to hand over.
An editable table. A 4B model will sometimes be wrong, so every cell is fixable. Change a vendor's category once and Receipt Box remembers it: that vendor is sorted the same way from then on.
Duplicate detection, so the same invoice uploaded twice gets flagged instead of counted twice.
Multi-currency handling, with a currency switch instead of silently adding euros to forints.
Exports to CSV and to an Excel workbook (receipts, spend per category, VAT per quarter) that goes straight to the accountant.
Optional line-item extraction for when you want to see what was bought, not just the total.

Who would pay? Freelancers and tiny agencies who want bookkeeping help but cannot (or should not) upload client invoices, travel documents and personal receipts to a cloud service. The same goes for the accountants who serve them: with many clients' paperwork on their desk, "nothing leaves this machine" is a feature they can promise. I see it as a one-time licence for freelancers, or a small monthly fee per accountant seat.

Demo

Live app: https://receipt-box-69v1.onrender.com (hosted on Render; click Load demo data to explore the dashboard)

The repo includes three fictional sample documents in samples/ (a PDF invoice and two receipt images in different currencies). Drop them in and the dashboard fills in as each document is read.

Code

banana0000 / receipt-box

Private receipt & invoice sorter for freelancers: Gemma 3 reads your receipts locally via Ollama. Editable table, learned categories, VAT by quarter, multi-currency, CSV/Excel export.

Receipt Box

A receipt and invoice sorter for freelancers that runs entirely on your laptop Photos and PDFs go to an open-weight vision model (Gemma 3 via Ollama), and the result is a categorised expense dashboard plus a CSV or Excel file for your accountant Nothing is uploaded anywhere.

Features

Reads receipt photos (JPG / PNG) and PDF invoices (text layer or scanned) with Gemma 3.
Dashboard: total spend, VAT to reclaim, spend by category and by month, VAT by quarter.
Editable table: fix any cell, delete rows. Changing a category teaches the app: that vendor is sorted the same way from now on (saved in rules.json).
Possible duplicates are highlighted.
Several currencies: pick one, or "All currencies" to see them side by side. Amounts are never converted.
Export to CSV or Excel (receipts, spend per category, VAT per quarter).
Optional line-item extraction (slower).
Files are read one at…

View on GitHub

How I Built It

The whole project is built around one open-weight model: Gemma 3 (4B, vision), running locally through Ollama. The only network call the app makes goes to localhost.

Structured output. Each image is sent to Gemma with a JSON schema (vendor, date, total, vat, currency, category), and the category is constrained to a fixed list. The reply always parses, with no regex and no retry loop.
Two paths for PDFs. If a PDF has a text layer, the text goes to the model directly, which is faster. Scanned PDFs are rendered to an image (PyMuPDF) and go through the vision path.
Queue with visible progress. On a CPU a receipt takes roughly 15-20 seconds, so files are processed one at a time and the page shows "Reading 2 of 5: name.png" instead of freezing.
The model is allowed to be wrong. The prompt tells it to use only what is printed and not to guess. After that, the human-in-the-loop table and the learned vendor rules do the rest. In my tests the three sample documents were read correctly, and for real-world crumpled receipts the correction table is what makes the tool usable.
Dashboard. A Dash app with Plotly charts and pandas for the numbers (quarterly VAT, duplicates, per-currency totals). Excel export uses openpyxl.
Tests. The data layer (duplicates, quarters, rules, exports) is covered with pytest.

Hosting. The same app also deploys to Render from a Blueprint (render.yaml): a private Ollama service with Gemma 3 baked into its Docker image, and the web app talking to it over Render's private network, so the model is never exposed to the internet. Gemma 3 4B needs about 4-5 GB of RAM, so the Ollama service runs on a large instance. On a laptop none of this is needed.

Honest limits: a 4B model on a CPU is slow, and line-item extraction is slower still, so it is off by default. A bigger Gemma or a GPU is a one-line change (RECEIPT_MODEL).

Why Does Open Innovation Matter?

Receipts are about the most sensitive "boring" data a freelancer has: who you work for, where you travel, what you pay for. A closed API would force a trade-off between convenience and privacy, and that is exactly where I would lose the customers who care about it most.

An open-weight model removes the trade-off. Because Gemma runs on the user's own machine:

the documents never leave the laptop, so there is nothing to leak, retain or train on;
there is no per-receipt API bill, so the price of the product is not tied to someone else's pricing;
it keeps working offline, on a train or in a client's office with no Wi-Fi;
the user can swap the model by changing one environment variable as better open models arrive.

A small open model that runs on a laptop is what makes "private by design" a real promise here, not a marketing line.

Prize Categories

Gemma: the entire product is built on Gemma 3 running locally.
Render: the web app and a private Gemma/Ollama service are deployed together from one Blueprint.

Exam AI – something I built during the MeDo hackathon

E. Smith — Mon, 04 May 2026 07:30:20 +0000

🚀 During the MeDo hackathon,I built something surprisingly simple — but genuinely useful: Exam AI 📚🤖

The idea came from a pretty common frustration.

A lot of studying turns into passive repetition:
reading the same notes again and again, scrolling slides, highlighting things… and hoping it somehow sticks.

I wanted to build something more active.
Something that actually makes you think while learning.

So I created Exam AI. This AI tool is very easy to use.

You give it a topic, and it:
✨ generates exam-style questions
✨ lets you answer interactively
✨ explains the answers clearly
✨ helps you dig deeper when something doesn’t fully click

It feels less like a quiz app and more like studying with someone who keeps asking the right questions.

⚙️ Building it

I used MeDo to build the project, which let me move really fast.

Instead of spending the whole hackathon dealing with setup or infrastructure, I could focus on:
🧠 refining prompts
🧠 improving explanations
🧠 making the experience actually helpful

Most of the work wasn’t technical complexity — it was figuring out how to make the AI respond in a way that feels useful and natural.

🧩 The difficult part

The biggest challenge was keeping things consistent.

With AI, small changes in prompts can completely change the quality of the output.
Some questions were great. Others… not so much 😅

We also had to constantly fight the urge to add more and more features.

In the end, keeping the product focused was probably one of the best decisions we made.

📚 Biggest takeaway

Hackathons really remind you that:
⚡ speed matters
⚡ simple ideas win
⚡ shipping something beats endlessly planning it

And honestly, AI becomes much more powerful when you focus less on “cool features” and more on helping people in a practical way.

I am excited to keep improving it — especially around personalization, progress tracking, and adapting to weak areas over time.

But for now, we’re just happy we turned an idea into something real 🙌

You can check my app here:
https://medo.dev/projects/app-bdybwhukvg8x
https://www.youtube.com/watch?v=aFDX8r6-ULY

🎃 My Kiroween Hackathon Experience — I Built a Spooky Landing Page with MCP Integration using Kiro IDE

E. Smith — Wed, 26 Nov 2025 10:17:39 +0000

I joined the Kiroween hackathon with one goal: build something fun.
A few late-night ideas later, I ended up creating a Halloween-themed, API-powered landing page — entirely by myself, inside Kiro IDE.

👻 What I made

Spooky Landing Page with MCP Integration — an animated, glowing UI that pulls real-time spooky quotes via MCP.
It reacts, flickers, whispers Halloween mood into the screen.

Features:

🔮 MCP + ZenQuotes API connection
💀 Animated UI states (ghost / moon / crystal ball)
🎨 DiceBear avatar generation
⚡ Smooth transitions + spooky sound cues

Short project — big vibes.

🧰 Built in Kiro IDE

I didn’t code this traditionally — I built it through conversation.

Using Kiro IDE, I generated components, refined effects, and integrated MCP endpoints quickly.
It felt less like typing code, more like designing an idea in real time.

🏁 Why this mattered to me

I learned how fast a solo developer can move when the tooling flows with creativity.
Kiro IDE helped me think, build, and refine — not just code.

Small project, but it left a big smile.👻🔥

DEV Community: E. Smith

Slothstep: a lazy-friendly app that turns "ugh, I'll do it later" into three tiny steps

What I Built

Demo

banana0000 / slothstep

Slothstep: turn a task you keep putting off into three tiny steps with an open-weight model (Gemma 3 + Ollama). Dash app.

🦥 Slothstep

Features

Run locally

Configuration

How I Built It

Why Open Innovation Matters Here

Prize Categories

Receipt Box: a private receipt sorter for freelancers, powered by Gemma

What I Built

Demo

Code

banana0000 / receipt-box

Private receipt & invoice sorter for freelancers: Gemma 3 reads your receipts locally via Ollama. Editable table, learned categories, VAT by quarter, multi-currency, CSV/Excel export.

Receipt Box

Features

How I Built It

Why Does Open Innovation Matter?

Prize Categories

Exam AI – something I built during the MeDo hackathon

⚙️ Building it

🧩 The difficult part

📚 Biggest takeaway

🎃 My Kiroween Hackathon Experience — I Built a Spooky Landing Page with MCP Integration using Kiro IDE

👻 What I made

🧰 Built in Kiro IDE

🏁 Why this mattered to me