Network anomaly detector, https://www.researchgate.net/publication/326699313_Network_Anomaly_Detector_using_Machine_Learning, 07/01/18, A research paper on an automatic network anomaly detector published on International Journal of Engineering & Technology on July 2018. Python Scraper Framework, https://github.com/RedGlare96/PythonScraperFramework, 09/01/20, A custom python Framework that integrates all latest techniques to automate interactions with web pages while bypassing most anti-bot detection systems implemented by the target sites. All the required steps are abstracted on a package but are built to be customizable to suit any use-case. This framework lays the groundwork to make a Python web scraper. A scraper for a single webpage can be created in minutes with just the inbuilt helper functions. You can also create your own custom scraping routines. It uses chromedriver headless by default but the method can be overridden to whatever method you require. Page Miner, https://github.com/RedGlare96/PageMiner, 10/01/20, A Python script that can crawl through any number of links in an entire website and scrape specific html tags in each individual page as per input configurations. Python scraper that scrapes specific tags from a website. Crawls to links within the main page and repeats the process n number of times as specified in the input. The tool can be used by a layperson to create advanced web crawlers with just a few inputs. Spotify Playlist Manager, https://github.com/RedGlare96/PlaylistManager, 01/01/21, A Python Flask web interface layer of the Spotify open api that allows the user to automate creating playlists, maintaining order of tracks in a playlist and managing existing playlists. The full list of features include: Login with your username and password, Flask Web interface, Can manage multiple playlists, Can manage existing or create new playlists, Randomly takes tracks from a random playlist when creating a new playlist, Set priority tracks that have a fixed position in the playlist, will revert back to set position even if moved. NDS Captive Portal, https://github.com/RedGlare96/NdsCaptivePortal, 08/01/22, A Django implementation of a Captive Portal: The webpage you receive when connecting to any public wifi. It supports: 1. Creation of accounts with a dedicated admin panel to allow staff manage them, 2. Setting custom browsing time limits for each account, 3. Tracking of account activity including total browsing time, log in times, etc, 4. Creation of discount vouchers with a user interface to redeem them and an admin panel for staff to create and manage them StatScraper, https://github.com/RedGlare96/StatScraper, 05/01/22, An automated scraper that can collect stats (likes, followers, subscribers, etc) from multiple social media websites. It can be used by a layperson who just needs to provide inputs on the sites and get a fully featured excel sheet with scraped stats for the entire list. Each site will have its own worksheet on the same excel document. It even has support for non-public pages, allowing you to login to accounts once and auto-logging in with the saved accounts until the cookies are valid. Currently supported sites: Youtube, Trustpilot, Twitter, Facebook, Tikitok GPT-3 Chatbot, 11/01/22, A GPT-3 chatbot service for a brand that can answer questions about the brand, its products, the programmes they offer, their pricing and many more. The bot is powered by OpenAi's GPT-3 functionality and uses the da-vinci model using fine tuned models. The bot can be interfaced via a web interface made in Flask, or via SMS which works using the Twilio service which can be leveraged to both send and receive SMS messages. The chatbot works completely autonomously hosted on a remote server and is able to engage and answer questions from multiple customers at once from multiple interfaces at the same time with no human intervention whatsoever. GPT-3 AI Assistant, 01/01/23, A GPT-3 powered AI assistant that uses OpenAI's completion model to create a language model that can convert commands typed in natural language to a form a backend can understand in the form of 3 finite instructions with changing parameters. These converted instructions are parsed by a Python backend which then carries out the specific task with the given parameters, following up with questions if some parameters are missing. Settl backend, https://thesettl.com/, 07/01/22, Settl is a website that acts as a portal for users seeking PG accommodation in India. They can view available rooms, look up amenities available, request a virtual tour, book a room and initiate a contract with the PG owner using the website or the app. I worked as a backend developer for the Settl website which was developed in Django. The things I worked on were: Fixing old issues with code, Working on the contract creation system, Developing api to be used by Android and iOS apps, integrating payment services (Cashfree and Razorpay), Managing webhooks, Managing deployment and performing Devops