DEV Community: Parag Ghatage

I’m building a Python-native frontend framework that runs in the browser

Parag Ghatage — Sat, 22 Nov 2025 09:15:20 +0000

For years, the browser belonged entirely to JavaScript.

I decided to challenge that assumption.

I’m currently building Evolve **- a **Python-native frontend framework powered by WebAssembly and a minimal JavaScript DOM kernel.

The goal is simple:

Write UI in Python

Run it in the browser

Keep it fast, reactive, and simple

I’m still deep in development, so I’m not publishing the source yet.

But I will be sharing progress, architecture, and demos.

If you’re curious about Python + WebAssembly in frontend, stay tuned.

One Month into GSoC with AOSSIE: My Journey Building Perspective

Parag Ghatage — Tue, 08 Jul 2025 20:11:14 +0000

🚀 TL;DR

I’m one month into Google Summer of Code with AOSSIE, building Perspective- Perspective analyzes your news or social feed and presents credible counter-narratives from reliable sources-helping you think critically, reduce bias, and see the full picture. Don’t settle for one-sided stories. Get complete, nuanced facts.

🎯 1. Project Overview

Perspective analyzes news articles to surface hidden biases by combining cleaning, entity extraction, and bias scoring. I’m collaborating with mentors Manav, Pranavi and Bruno on the Perspective repo.

📆 2. Month 1 Milestones

Week	Deliverable	Status
1	Project setup & Figma mockups	✅ Done
2	Frontend pages (Landing, Loading, Results)	✅ Merged PRs
3	`cleaner.py` & `extractor.py` modules	✅ Merged PRs
4	UV‑based backend, scraper module, LangGraph	✅ Merged PRs

📔 3. Day‑by‑Day Highlights

June 2–5:

– Revamped frontend: Landing & Analyze pages in Next.js+Tailwind.

– Built Loading & Results screens with global state flow.

– UI/UX tuning and responsiveness fixes after mentor feedback.
June 6–10:

– Squashed a React‑date‑picker build bug.

– Finalized frontend deployment on Vercel.

– Explored uv for Python dependency management; set up new backend architecture.

– Added a FastAPI scraper module for article ingestion.
June 11–15:

– Dug into LangGraph: designed the pipeline flow (Sentiment → Fact‑Check → Generate → Judge → Retry).

– Implemented base StateGraph and stubbed out node files (sentiment.py, fact_check.py, etc.).

– Researched error‑handling patterns and added an ErrorCatcher node.
June 16–20:

– Integrated Groq’s Python SDK for sentiment analysis; tuned prompts for deterministic outputs.

– Built DuckDuckGo search node and LLM‑analysis node for real‑time fact‑checking.

– Crafted structured CoT prompts and wired up the generate_perspective node.
June 21–25:

– Developed & integrated judge_perspective scoring node (originality, reasoning, factual grounding).

– End‑to‑end tests on real articles (e.g., 2025 French Open final).

– Refactored for stability: added fallbacks, cleaned up exception flows.
June 26–30:

– Chunking utility for vector DB ingestion; set up Pinecone embeddings & metadata storage.

– Linked frontend to backend for fact‑check results; switched from Render to Hugging Face Spaces for deployment.

– Automated backend CI/CD with GitHub Actions; local testing via act.
July 1–7:

– Polished docs & example pipeline states; pushed final PRs.

– Monitored deployments, fixed minor bugs, and synced progress with mentors.

🛠 3. Deep Dive: What I Built

3.1 LangGraph Workflow

Challenge: While building LangGraph, efficiently handle all the data and making sure to pass right state takes lot of efforts.
Key Learnings: LangGraph workflow needs logging state after each node and robust error handling.

🤯 4. Challenges & Learnings

LangGraph quirks:
- I hit a bug when chaining async nodes. Fixed by adding explicit await and restructuring my graph definition.
- Tip: Write unit tests for each node in isolation—saved hours of debugging.
uv in deployment :
- uv needs carefull docker setup for deployment.
- Fix: loaded uv in container and ran my main.py with --no-cache paramenter to prevent container permission issues.
Collaboration flow:
- Coordinating with three mentors across different time zones taught me to write crystal‑clear PR descriptions & use GitHub Projects for tracking.

🚀 5. What’s Next (Month 2)

Bias Scoring Module: Prototype a heuristic‑based scorer combining sentiment analysis + entity context.
Vector DB Integration: Push cleaned articles & metadata into Pinecone for fast similarity queries.

I’ll be posting updates every 2–3 weeks-stay tuned!

🙏 Call to Action

- Feedback welcome: Spot a bug, have a shortcut, or a study tip? Drop a comment below.

“The journey of a thousand miles begins with a single PR.” – me, off to write my next one 😅

Thanks for reading!

Parag Ghatage

GitHub · X · Portfolio
Linkedin

Artify: Leveraging Deep Learning to Transform Images into Unique Art Pieces

Parag Ghatage — Sun, 10 Nov 2024 14:52:37 +0000

Try Artify Here!

In the age of digital creativity, image editing, and artistic transformations, “Artify” is an innovative leap toward the future of AI-driven design. This project combines my passion for machine learning and creative technology to produce a platform where users can effortlessly transform ordinary images into artistic masterpieces. Here’s an inside look at how Artify was conceptualized, the tech stack behind it, and the impact it aims to make.

What Is Artify?
Artify is an AI-powered web app that turns images into customized, AI-generated artwork in seconds. By leveraging deep learning techniques, the app interprets user-uploaded images and applies unique styles, colors, and textures, giving each image a distinctive and eye-catching flair. Whether it’s turning photos into watercolor paintings or giving them a modern art twist, Artify provides users with a flexible, easy-to-use platform to explore their artistic side.

The Inspiration Behind Artify
I have always been fascinated by the way AI can be used to bridge technical and creative domains. The idea of Artify came from my interest in image processing, neural networks, and the desire to make art accessible to everyone. I wanted to create something that could inspire people, including those who may not have a background in art, to see their photos in a whole new light and to recognize the artistic potential that machine learning holds.

Key Features of Artify
Customizable Art Styles: Users can choose from a wide range of art styles such as oil painting, pencil sketch, watercolor, and abstract art. Each style has its unique set of parameters that the user can fine-tune.
Real-Time Preview: Users can see the transformation as it happens, allowing them to adjust the style parameters and see the results instantly.
High-Resolution Outputs: Artify doesn’t compromise on quality. The AI models are optimized to produce high-resolution images, making them suitable for printing or digital portfolios.
Simple UI and UX: One of my goals was to ensure Artify is intuitive. Even users without a tech background can use Artify with ease, transforming their photos into stunning art with just a few clicks.
Cloud-Hosted Processing: Artify is deployed on Google Cloud Run, which allows it to handle a high volume of requests while maintaining performance. Cloud hosting also makes it accessible from anywhere in the world without the need for heavy local resources.
The Technology Stack
The backend of Artify relies on machine learning models specifically trained for image stylization. Here’s a breakdown of the tech stack that powers Artify:

Frontend: Built with Next.js, the frontend provides a responsive and engaging user experience, making the app accessible across devices.
Backend: The backend is developed in Flask, handling requests from the frontend and routing them to the machine learning model.
Machine Learning Model: Artify uses a combination of TensorFlow and Keras, where I applied a neural style transfer algorithm. This model is trained on various datasets to generate different artistic styles.
Cloud Deployment: Artify is hosted on Google Cloud Run, ensuring scalability, reliability, and security. Cloud Run also provides automatic scaling based on the number of requests, making it an ideal choice for handling high demand.
How Artify Works Under the Hood
Artify’s core functionality revolves around a neural style transfer model, which applies artistic patterns and textures to images. Here’s a simplified view of the process:

Image Upload: Users upload an image to Artify, which is processed in real-time on the server.

Style Selection and Parameter Tuning: The user selects a style and adjusts parameters like brush size, texture density, or color saturation. These parameters tweak the model’s output, allowing for unique results each time.

Neural Style Transfer: The backend model extracts the content from the uploaded image and the stylistic features from the chosen art style. It then blends the two to create an entirely new image with the content of the original and the visual characteristics of the selected style.
Image Generation and Download: Once the image is processed, users can download the high-resolution output for printing or digital use.

Generated image with painting style applied
Challenges Faced and Lessons Learned
Developing Artify wasn’t without challenges. One of the main issues was ensuring that the model could handle high-resolution images without taking too long to process them. Training the model on a diverse dataset helped, but optimizing performance for real-time output took considerable effort.

Another challenge was maintaining quality while deploying on the cloud. Google Cloud Run helped streamline this with its auto-scaling features, but testing and setting up the environment to ensure smooth performance and manage costs was a balancing act.

Through Artify, I learned invaluable lessons about optimizing machine learning models for production, managing cloud resources, and creating user-centric applications that blend art and technology.

The Future of Artify
Artify is a project that I believe has room to grow. Here are a few features I’m considering for future iterations:

More Styles and Effects: Adding more unique styles, such as pop art, pointillism, and manga-inspired effects.
AI-Based Style Recommendation: An AI feature that suggests art styles based on the image’s content.
Social Sharing and Community: Allowing users to share their creations on social media directly from the app. Eventually, building a community where users can share and discover art created through Artify.
Batch Processing: Enabling users to upload multiple images at once and apply the same effect, ideal for creators looking to process large collections for portfolios or projects.
Conclusion
Artify represents more than just an art transformation tool. It’s a step towards making AI-driven creativity accessible, easy, and inspiring. With its powerful features, simple interface, and unique results, Artify is a testament to the endless possibilities of merging technology with art.

I am excited to see how Artify evolves and hope that it serves as a source of inspiration for artists, creators, and technology enthusiasts alike. Whether you’re looking to experiment with digital art or want to reimagine your favorite photos, Artify is here to turn your images into something extraordinary.