Document Summarizer

piyushmishra06 — Mon, 26 Oct 2026 06:32:23 +0000

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

What I Built

Documents are always hectic to read , often they have their details tucked. A document Summarizer helps ease the tedious task of reading documents and understanding the key parts of it. I built docAnalyzer for that.

Code

https://github.com/piyushmishra0620/HacktoberFest-Week-3

How I Built It

Used fastApi backend to gathers the form data from the frontend through /api/Summarizer , extractor worker gets into work , picks the document to be read the document , scans them thoroughly and extracts the text . It uses the chunking method where the text longer than 12k characters are broken down into segments . It then sends the request to locally running gemma , which Summarizes the entire thing and gives out the response .
The response is served to the frontend.

The Lazy Payoff

It saves a lot of time as a reward , where people with less effor can be able to know all about the document , rather than reading it from top to bottom in a detailed manner.

MatchForge - Resume Analysis And Job Search Tool

piyushmishra06 — Mon, 19 Oct 2026 06:54:28 +0000

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

What I Built

A Resume analysis and Job Search tool where users can upload their resume , get their ATS score for the job which they are applying and finally get some recommendations about areas of improvement in the resume along with the matching jobs from across various sources. People often don't know the exact things they should write inorder to get a higher chance of selection besides they apply at places they can't hope much to get selected . They need the right companies where they will gain edge at, not the ones where they may get rejected . So, this task is a tedious one : to find the things missing in their resume or rectify some , and then to get the idea of the correct places to apply at . MatchForge makes it simpler . Just upload the resume , give the title of the job for which you are applying , then it will give the ATS score along with areas of improvements and then the list of places at which they can apply at . This will atleast give people an idea of the type of companies they should be aiming for.

Demo

-> Here's the drive link showing the demo to use it : https://drive.google.com/file/d/1BTS_ShTyeeo8oAPvmWOSovR80NwNXzcy/view?usp=sharing .

Code

-> The link to source code : https://github.com/piyushmishra0620/Hacktoberfest-Week-2 .

How I Built It

I made a react application with an expressjs server , where users upload their resume and the role for which they are applying . Afterwards, the resume text is extracted and sent along with the job description to the locally running model - gemma3 in ollama . It analyzes all the things in resume , matches with the jobs thereby returning with the resume score , the scope of improvements and matching jobs . It takes upto 2 minutes to display the entire thing in the frontend . With more power and GPUs , time can be greatly reduced for the model to perform better .

Prize Categories

-> I made use of gemma3 . Hope this is worthwhile to pay to use it!

DEV Community: piyushmishra06

Document Summarizer

What I Built

Code

How I Built It

The Lazy Payoff

MatchForge - Resume Analysis And Job Search Tool

What I Built

Demo

Code

How I Built It

Prize Categories