DEV Community: Mohammad Sami shaikh

NeuroPulse a ai/ml chatbot

Mohammad Sami shaikh — Wed, 21 Oct 2026 10:38:37 +0000

NeuroPulse: The Sloth's Guide to Mastering AI/ML with 100% Local Open-Source AI

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

What I Built

Keeping up with Artificial Intelligence and Machine Learning feels like a full-time job. Between 80-page whitepapers, convoluted mathematical equations, 3-hour video tutorials, and endless forum debates, understanding even basic concepts like backpropagation, contrastive loss, or transformer attention usually takes hours of digging.

In honor of International Sloth Day and this week’s "Slow and Steady" challenge, I built NeuroPulse — a local, privacy-first Retrieval-Augmented Generation (RAG) assistant designed to reward intellectual laziness.

Instead of wading through dozens of open browser tabs, reading bloated medium articles, or paying monthly subscription fees to proprietary AI APIs, NeuroPulse does all the heavy reading for you. You ask a direct question, and it retrieves grounded knowledge from a curated vector database, validates the topic, and synthesizes a punchy, accurate answer with verifiable source citations — running 100% offline on your own machine.

Developers, students, and engineers who dread information overload get to be a whole lot lazier without falling behind.

Demo

NeuroPulse features both a clean, modern conversational web interface and a terminal CLI:

![NeuroPulse Chatbot Interface]

Code

The complete source code is open source under the MIT License:
https://github.com/Mohammad1sami/AI-ML-chatbot

Project Architecture at a Glance

AI Chatbot/
├── aiml_knowledge_base.csv       # 118 verified AI/ML knowledge entries & categories
├── chroma_langchain_db/          # Persistent local Chroma vector database
├── chroma_langchain_db.csv.sha256# SHA-256 change detector for zero-waste indexing
├── vector.py                     # Chunking, mxbai embeddings, and Chroma retriever setup
├── server.py                     # FastAPI server with SSE streaming & REST endpoints
├── main.py                       # CLI interactive runner
└── static/                       # Fast, minimalist frontend (HTML5/CSS3/Vanilla JS)
    ├── index.html
    ├── style.css
    └── app.js

How I Built It

NeuroPulse is built exclusively on top of open-source AI and open-weight models — no closed cloud endpoints or commercial APIs:

1. The Open-Weight AI Stack

Open-Weight Reasoning & Generation: Meta's Llama 3.2 (llama3.2), executed locally via Ollama. Llama 3.2 provides exceptional reasoning and conversational capabilities while running comfortably on consumer laptop hardware.
Open-Weight Embeddings: mxbai-embed-large (Mixedbread AI via Ollama). A top-tier, high-dimensional open embedding model tailored for dense retrieval tasks.
Open-Source Vector Database: ChromaDB (langchain-chroma), persisted directly to disk in ./chroma_langchain_db.
RAG Orchestration: LangChain (langchain-ollama, langchain-text-splitters, langchain-core) chaining the prompt templates, topic classification, and document retrieval.
Backend API: FastAPI & Uvicorn with asynchronous Server-Sent Events (SSE) streaming and REST endpoints.

2. Dual-Stage Guardrail Pipeline

To keep things fast and prevent hallucinated ramblings, the system implements two LangChain chains:

Topic Classifier Chain: Uses llama3.2 to classify whether the incoming query is genuine AI/ML. Unrelated questions (sports, celebrity gossip, recipes) are politely deflected before doing expensive vector math.
Context-Grounded Answer Chain: Constrains the model strictly to verified context chunks retrieved from ChromaDB, ensuring direct and accurate answers.

# Automatic SHA-256 change detection: only re-index when your knowledge base edits!
with open("aiml_knowledge_base.csv", "rb") as f:
    csv_hash = hashlib.sha256(f.read()).hexdigest()

if csv_hash != previous_hash:
    # Split, embed with mxbai-embed-large, and persist into ChromaDB
    chunks = splitter.split_documents(source_documents)
    vector_store.add_documents(documents=chunks, ids=ids)
    with open(HASH_PATH, "w", encoding="utf-8") as f:
        f.write(csv_hash)

Why Does Open Innovation Matter?

Open innovation is what makes this project practical, trustworthy, and sustainable:

Zero Cost & Infinite Free Queries: Proprietary LLM APIs charge per token. A developer querying an API 200 times a day faces monthly bills and credit card limits. With open-weight models via Ollama and ChromaDB, the marginal cost per query is exactly $0.00.
100% Offline & Absolute Privacy: Many developers and researchers study proprietary algorithms or confidential datasets. Because NeuroPulse runs completely on localhost, no prompt, question, or document ever leaves the device. It functions seamlessly on an airplane, in a remote cabin, or behind strict enterprise firewalls with zero internet connection.
Auditability & Explainability: Closed AI systems are black boxes that frequently change their behavior without warning. With an open CSV dataset, open embeddings, and inspectable ChromaDB vectors, every answer points back to a transparent source topic and row ID.
No Vendor Lock-In: If a newer open-weight model drops tomorrow (like Mistral, DeepSeek, or Gemma), swapping it in takes a single line change in server.py (model = OllamaLLM(model="new-model")).

The Lazy Payoff

Here is the sloth math on how much effort NeuroPulse saves:

Searching 15 Browser Tabs vs. 1 Query: Traditional research for a concept like L1 vs. L2 Regularization requires searching Google, skipping sponsored ads, dodging cookie popups, and parsing multiple contradictory StackOverflow answers (typically 12–18 minutes). NeuroPulse delivers the synthesized definition and mathematical trade-off in under 3 seconds.
Automated SHA-256 Knowledge Maintenance: You never have to manually rebuild vector indexes or run maintenance scripts. The lazy system checks file hashes on startup — if the CSV hasn't changed, it boots instantaneously using the cached index.
Estimated Savings: Over a week of studying machine learning topics, asking 10 questions a day saves approximately 2 hours of reading fluff and tab clutter.

Prize Categories

Open-Source AI / Open Weights Track (Powered by Ollama, Llama 3.2, mxbai-embed-large, ChromaDB, and LangChain)
Slow and Steady (Week 3 Challenge)

Built with ❤️ for Hacktoberfest 2026. Happy Sloth Day — work less, build open, and let the local models do the heavy lifting!

BruteWing: The Offline-First AI Gym & Client Ledger Built for Local Fitness Hustles

Mohammad Sami shaikh — Wed, 14 Oct 2026 16:13:20 +0000

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

What I Built

The Problem: Why Small Fitness Studios Hate Gym SaaS

If you talk to independent gym owners, CrossFit box coaches, or freelance personal trainers running their own training spaces, you will hear the exact same complaint: commercial gym management platforms are bloated, overpriced, and fundamentally broken for local operations.

Platforms like Mindbody or Zen Planner charge $80 to $250+ every single month. For a trainer with 30 to 100 members, that recurring SaaS bill eats a massive portion of their margin. Worse:

Basement gyms and warehouse boxes are cellular dead zones. When the Wi-Fi drops or cell service is spotty, cloud-dependent SaaS apps freeze or refuse to load member check-ins.
Client privacy is compromised. Confidential member notes (injuries, body composition, dietary restrictions, phone numbers, and payment details) are harvested onto commercial cloud servers.
Trainers fall back to chaotic notebooks and WhatsApp chats. When software gets too complicated, trainers abandon it for paper ledgers or loose messaging threads, forgetting renewal dates and losing thousands in uncollected fees.

The Solution: BruteWing (GymLedger)

BruteWing is a clean, local-first gym management ledger and training assistant built with Flutter and powered by an open-weight AI model running on-device.

It solves the operational headaches of running a fitness studio without requiring recurring subscriptions, cloud lock-in, or even an active internet connection.

Key Features:

📊 Dynamic Membership & Expiry Cockpit: Real-time status cards displaying active, expiring (≤ 5 days), and expired memberships with visual day-count countdown rings.
💳 Complete Billing & Payment Ledger: Log and manage monthly, quarterly, and annual plans across Cash, UPI, Card, and NetBanking. Filter by status (Paid, Pending, Overdue).
🤖 Local AI Diet & Workout Synthesis: Powered by an open-weight LLM running on-device (via local inference). It turns quick coach notes (e.g., "1900 kcal cutting, anterior knee pain, wants 4-day split") into structured progressive workout splits and macronutrient breakdowns with zero cloud API latency.
📲 1-Click WhatsApp Follow-ups: Automatically crafts personalized renewal nudges and payment receipts, opening directly into WhatsApp with dynamic parameters ready to send.
🔒 100% Offline-First SQLite Architecture: All data resides in encrypted local storage (sqflite / shared_preferences). It works flawlessly in basements, steel-roof gyms, and remote training camps.

Who Would Pay For It?

Independent gym owners & studio managers looking to replace $100+/month software with an affordable one-time purchase or low-cost self-hosted utility ($29–$49 / ₹999–₹1,999).
Freelance personal trainers & CrossFit coaches who need a lightweight, professional client tracking tool that fits in their pocket and works without Wi-Fi.

Demo

Here is a glimpse of BruteWing in action:

Studio Dashboard	Member Directory & Renewal Alert
Visual ring indicators for active vs. expiring memberships & monthly revenue overview	Color-coded status pills (`Active`, `Expiring`, `Expired`) with quick filter tabs

Payment & WhatsApp Reminder	Local AI Training & Diet Notes
1-tap WhatsApp reminder dispatcher with auto-filled polite nudges & transaction tracking	Personalized nutrition targets and periodized workouts synthesized locally

Check out the live repository or build artifact:

GitHub Repository: https://github.com/Mohammad1sami/gym_ledger

- Video Link: https://drive.google.com/file/d/17Hd1I4EPDYrQpkCxD8mesIMcLSksA0jL/view?usp=drivesdk

Code

The application is written in Flutter (Dart) using a modular architecture with SQLite persistence and local inference hooks.

# AniMix 🌿 — Touch Grass, Photograph Real Animals & Fuse Them into AI Monster Cards!

Mohammad Sami shaikh — Fri, 09 Oct 2026 09:57:11 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built

AniMix is an offline-first, client-side web application built to get gamers and tech enthusiasts off their screens, out of their rooms, and into the wild.

The concept is simple: Touch Grass. Photograph Animals. Fuse Monsters.

🌿 How It Works:

Explore the Outdoors: Venture outside into your yard, neighborhood, or local hiking trail to search for real living fauna—dogs, cats, birds, squirrels, butterflies, or bugs.
On-Device Vision AI: Snap a photo using your smartphone or laptop camera. AniMix runs an open-weight vision classification model directly inside your browser to identify the animal species instantaneously.
Monster Fusion: Select any two captured animals from your deck (for instance, a Golden Retriever + a Blue Jay) and fuse them together! AniMix procedurally calculates combat stats, element affinities (Bio, Aero, Pyro, etc.), and generates a custom hybrid monster collectible card.
Collect & Share: Export high-resolution holographic trading cards to build your creature deck or share them with friends via the Web Share API.

Demo

🌐 Live Web Application: https://animix-touchgrass.vercel.app
🎬 GitHub Repo: https://github.com/Mohammad1sami/animix

Code

https://github.com/Mohammad1sami/animix

How I Built It

AniMix is built around client-side open-source AI and modern web technologies:

Open-Source AI & Local Inference: Powered by Transformers.js using the open-weight MobileViT-small model (Xenova/mobilevit-small). The model runs entirely in-browser via WebAssembly and ONNX Runtime, performing vision classification without any backend API call.
Frontend Framework: Built with Next.js 16 (React 19) for fast server-rendered shells and optimized client hydration.
Styling & Aesthetics: Styled using Tailwind CSS v4 featuring dark mode, glassmorphism, and dynamic animations.
Card Rendering & Export: Utilizes the HTML5 Canvas API and html-to-image for generating high-DPI monster trading cards with procedurally assigned stats based on classification confidence scores.

Why Does Open Innovation Matter?

Open innovation is what made AniMix possible. Here is why open-weight models and client-side inference matter:

100% Offline Wilderness Usability: When hiking in remote parks or nature trails with zero cell signal, traditional cloud vision APIs fail completely. Open weights running locally via WebAssembly allow AniMix to function anywhere off-grid.
Absolute User Privacy: Outdoor photos of surroundings stay on the user's device. No images are ever uploaded to third-party servers.
Zero API Cost & Unlimited Scalability: Proprietary AI APIs charge per request. Open innovation enables unlimited real-time image scans for $0 in server costs.

🌐 Community Wisdom

When building browser-based AI applications with Transformers.js, the developer community on DEV has shared key insights on WebAssembly performance and model caching:

Running Hugging Face Transformers in the Browser with Transformers.js by @xenova

Prize Categories

Overall Winner
Best Use of Open-Source AI / Local Inference

Bilal Masala POS App

Mohammad Sami shaikh — Mon, 05 Oct 2026 05:50:12 +0000

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend

Bilal Masala POS 🌶️

What I Built

I built Bilal Masala POS — a fast, intuitive, and modern Point of Sale (POS) mobile application designed specifically for my good friend Bilal, who operates a local spice and masala store.

The Problem

Running a spice shop comes with unique retail challenges. Spices are sold in multiple fixed package sizes (50g, 100g, 250g, 500g, 1kg) as well as custom weight amounts requested on the spot by customers. During peak hours:

Calculating prices manually caused long queues and customer wait times.
Frequent manual calculation errors occurred during multi-item orders.
There was no quick way to hand out clean digital or printed receipts.

The Solution

Bilal Masala POS streamlines Bilal's daily store operations by offering:

Variant-Based Product Management: Pre-configured spice products with variant weight options and prices.
Custom Weight & Manual Entry: Instant calculation for custom weights directly from the billing screen.
Real-Time Cloud Synchronization: Built on Firebase Cloud Firestore for instant inventory updates and category filtering.
Thermal Receipt & PDF Export: Instant 80mm thermal receipt generation with print and share options via WhatsApp or PDF.

Demo

🎥 Watch the Demo Video: Google Drive Demo Video

Code

Mohammad1sami / Bilal-masala-POS

bilal_masala

A new Flutter project.

Getting Started

This project is a starting point for a Flutter application.

A few resources to get you started if this is your first Flutter project:

For help getting started with Flutter development, view the online documentation, which offers tutorials, samples, guidance on mobile development, and a full API reference.

TODO // Add INC-DEC on onTap in Billing Screen // Remove Item from Billing Screen // add clear Items at clear button

View on GitHub

The complete source code is available on GitHub:
👉 GitHub Repository - Bilal-masala-POS

Tech Stack & Architecture

Frontend Framework: Flutter (Dart)
State Management & Routing: GetX
Database & Cloud Backend: Firebase Cloud Firestore
Receipt & PDF Generation: pdf and printing packages (customized for 80mm thermal receipt format)

How I Built It

I developed Bilal Masala POS using Flutter, GetX, and Firebase with the assistance of the Google Antigravity AI Agent.

By leveraging AI agentic capabilities:

Clean Architecture: Designed decoupled GetX controllers (HomeController, BillingController) and services (DatabaseService, PdfService).
Custom UI/UX: Created customized bottom sheets and reactive product cards tailored for rapid counter sales.
Thermal Receipt Layout: Built pixel-perfect 80mm receipt printing sections (Header, Customer Info, Items List, Totals, and Footer).

Why Does Open Innovation Matter?

Open innovation and open-source software level the playing field for small local business owners. Commercial POS systems are often bloated, expensive, and inflexible for traditional spice markets.

Thanks to open-source software like Flutter, community Dart packages, and open AI models, I was able to build a tailored, zero-subscription POS system for Bilal in just a few days.

My Agent Session

Developed with the assistance of an interactive Ollama AI Agent session, streamlining repository structure, state binding, and file untracking for open-source publication.

Prize Categories

Build for a Friend (Primary Category)