DEV Community

Bhavna Makode
Bhavna Makode

Posted on

LapWise – AI Laptop Advisor for Finding the Best Laptop

#devchallenge #weekendchallenge #hf26challenge #ai

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

What I Built

Buying a laptop can be overwhelming—between comparing specs across dozens of models, checking prices across multiple online retailers, and decoding technical jargon, my friend was struggling to figure out which laptop offered the absolute best value for their budget and specific needs.

To help them out, I built LapWise—an AI-powered laptop advisor.

LaptopWise automatically:

  • Scrapes and aggregates laptop hardware specifications and pricing from multiple tech storefronts.
  • Cross-references real-time market prices to surface the best available deals.
  • Analyzes my friend's exact budget, performance needs, and usage preferences using an open-source AI model.
  • Generates personalized recommendations with clear, plain-language explanations of why a specific model is the perfect fit.

Code

https://github.com/Bhavna2003/lapwise

Project Preview

How I Built It

LaptopWise was built using a modern full-stack setup focused around open-source AI and automated data collection:

  1. Open-Source AI Model: Powered by Ollama running Llama 3 / Mistral locally (or deployed via Hugging Face Inference API). The model acts as the decision engine, taking structured user requirements (e.g., "programming & light gaming, under $1000, good battery life") and matching them against normalized spec sheets.

  2. Data Aggregation & Price Comparison: Built a scraping/ingestion service in Python using BeautifulSoup and Playwright to extract price and spec updates across different e-commerce stores.

  3. Framework & Orchestration: Developed with LangChain / LlamaIndex to handle context retrieval (RAG) over the aggregated specs database, combined with a Next.js frontend for an intuitive conversational UI.

Why Does Open Innovation Matter?

Building LaptopWise around open-source AI and open tooling made a massive difference:

  • Complete Data Transparency: Closed commercial AI models can be biased or influenced by sponsored results. Using an open-weight model lets us audit the reasoning logic to ensure recommendations are 100% unbiased and truly best for the user.
  • Custom Logic & Fine-Tuning: Open weights allow us to customize system prompts, fine-tune context windows, and run fast local inference without worrying about strict API rate limits or recurring API fees.
  • Community Control: Hardware specs and deals change daily. Open innovation enables the community to contribute new store parsers, comparison metrics, and prompt strategies freely.

Top comments (0)