The AI-Powered Job Scraper and Notification System automates the job search process by extracting job listings and descriptions from platforms like LinkedIn and Indeed. It uses Natural Language Processing (NLP) models for job classification and skill extraction, providing personalized job recommendations based on predefined criteria. The system is built on a modular multi-agent architecture for web scraping, data cleaning, job classification, and notifications.
This project demonstrates the power of AI-driven automation in job searching by integrating web scraping, data processing, and NLP-based job recommendations in a scalable, efficient system.
Associated with Seattle University and Fortune 500 Company
Project Duration: Jan 2024 - Present
Algorithm for Fed Markets: This project, conducted in collaboration with a Fortune 500 company and affiliated with Seattle University, is a comprehensive endeavor aimed at revolutionizing market research strategies for a leading IT distributor. Focused specifically within the Space Force sector, the project introduces an innovative advanced search engine powered by cutting-edge unsupervised Natural Language Processing (NLP) techniques.
The primary goal of this initiative is to overhaul the distributor's internal data analysis and discovery process, which currently relies on manual research and data aggregation from disparate sources. By leveraging advanced NLP algorithms, the project streamlines the extraction, aggregation, and transformation of data, significantly enhancing efficiency and accessibility for the distributor's team.
Through the integration of a streamlined user interface and a robust data pipeline, the project enables seamless navigation through vast amounts of data from various sources. This automation not only optimizes the distributor's ability to identify emerging market trends and contracting opportunities but also empowers them to make well-informed decisions swiftly and stay competitive in the corresponding landscape.
Throughout the project timeline, spanning from January 2024 to June 2024, the team will collaborate closely with the Fortune 500 company and Seattle University to develop and implement state-of-the-art solutions tailored to the distributor's specific needs. By the project's completion, it is anticipated that the distributor will experience a significant reduction in both time and effort required for market research activities, ultimately leading to improved operational efficiency and strategic decision-making.
Can't Reveal much here because we signed NDA for the company so but can showcase our work regarding what we worked on : "ADVANCED SEARCH TOOL(NLP-Powered)", "Topic Modelling", "Name Entity Recognition", "Sentiment Analysis", "PowerBi Dashboard"
Project Duration: March 2024 - Present
Technical Synopsis: Streamlit-Powered ATS Resume Matcher with OpenAI GPT-3.5 Turbo
This project leverages OpenAI's powerful GPT-3.5 Turbo language model, seamlessly integrated within a user-friendly Streamlit web app. It utilizes CountVectorizer from Scikit-learn to transform resumes and job descriptions into numerical vectors for calculating cosine similarity and matching suitability. The system can dynamically generate job descriptions for different experience levels (entry, mid, and advanced) and suggest missing keywords to enhance resumes based on the job description, powered by GPT-3.5 Turbo. PyPDF2 library enables processing of PDF resumes, while secure key management safeguards API credentials. This Streamlit-based application significantly optimizes the resume screening process, offering a data-driven and user-friendly approach to talent acquisition for recruiters.
General Analysis complete!
| Technology | Entry-level | Mid-senior level | Advanced level |
|---|---|---|---|
| Computer Science | 47.41% | 53.78% | 57.07% |
| Data Science | 67.49% | 60.96% | 65.98% |
| Machine Learning | 55.56% | 39.30% | 46.02% |
| Business Analysis | 56.63% | 48.57% | 57.60% |
| Cloud Computing | 43.41% | 43.13% | 41.37% |
Skills: Large Language Models (LLM), Natural Language Processing (NLP), Application Programming Interfaces (API), Streamlit
Project Duration: September 2022 - December 2022
Technical Synopsis: Sound Recognition with Deep Learning for Bird Species Classification
This project explores the application of machine learning and deep learning techniques for the categorization of bird sounds, aiming to monitor bird population health and biodiversity. It covers the process of sound recognition from feature extraction to classification, utilizing spectrograms and neural networks. The methodology involves constructing a custom neural network for bird sound classification, including data preprocessing, binary and multi-class classification models, and transfer learning. Optimization algorithms and data augmentation techniques are employed to enhance model accuracy and robustness.
The project discusses the theoretical background of sound recognition, including feature extraction and classification using spectrograms and neural networks. It also covers hyperparameter tuning and optimization techniques to improve model performance.
This study successfully builds custom neural network models for bird species classification, explores transfer learning, and addresses challenges such as dataset size and overfitting. The report highlights the importance of sound recognition in monitoring bird populations and biodiversity.
Project Duration: May 2021 - Jun 2021
This project delves into the intricate behavior of stock markets on an annual basis, focusing on various disciplines such as NFLX, TSLA, and more. It addresses a wide array of challenges related to stock market analysis, including:
These techniques are complemented by effective visualizations using Seaborn and Matplotlib, providing insightful graphical representations of the analyzed data.
By working on this project, you will gain valuable skills and experience that are essential for becoming a proficient data engineer, analyst, and scientist:
Project Duration: Final year of Bachelor's project: Jan 2022 - Jun 2022
Data analysis involves understanding and interpreting a dataset to find answers to questions. Web scraping and visualization are powerful methods for automatically generating content on the internet. This project focuses on creating a movie rating forecast by extracting data from the IMDB website, enabling users to make informed decisions about which movies to watch based on ratings.
The project aims to develop an API that extracts data from multiple websites, preprocesses it, and visualizes it to provide business insights across various disciplines. By incorporating different perspectives into problem-solving, the project seeks to offer comprehensive solutions to queries.
The project utilizes Selenium and Beautiful Soup for web scraping, allowing users to extract data from different websites. The scraping process involves opening the website, inspecting HTML tags, and storing the desired elements into a data frame. Data scraping can be time-consuming and may require additional cleaning steps.
Web scraping enables the extraction of hidden web data, which is crucial for various applications. The project aims to provide an easy-to-use interface for searching, analyzing, and extracting data from websites. Future work may involve integrating machine learning techniques to automate decision-making processes.
Python's popularity in data science continues to grow, and future enhancements may include integrating machine learning algorithms to automate decision-making processes. Addressing the challenges of web structure inconsistency and implementing AI-driven applications are potential areas for future development.
Embark on a voyage into the world of data exploration with my first-ever project: Exploratory Data Analysis (EDA) on the Titanic Dataset. This project holds a special place in my heart as it marks the inception of my journey into the captivating realm of data science and machine learning.
As an aspiring data enthusiast during my undergraduate days, I embarked on this endeavor fueled by sheer passion and determination. Without the guidance of professors or faculty, I immersed myself in a sea of online resources, from courses on Coursera to tutorials on Udemy, diligently honing my skills and expanding my knowledge base.
This project was a pivotal milestone in my learning journey, intricately woven into the fabric of my pursuit of the Google Data Analytics Professional Certificate. Over the course of four months, I dedicated countless hours to mastering the intricacies of data analysis, drawing inspiration from every challenge encountered.
Initially, navigating through the complexities of data-driven insights seemed like an insurmountable task. However, with perseverance and unwavering determination, I gradually unearthed profound insights from the Titanic Dataset, unraveling the mysteries hidden within.
From discerning the distribution among passengers to uncovering correlations and factors influencing survival probabilities, every visualization and analysis served as a stepping stone in my evolution as a data scientist.
My journey extended beyond static datasets as I delved into the realms of web scraping, visualization techniques, and eventually, model training and optimization. Each step forward brought with it a deeper understanding of the intricacies of data science and machine learning.
As I delved deeper into model training and building, the culmination of my efforts manifested in the form of consistently accurate predictions, a testament to the efficacy of my methodology and the depth of my understanding.
More than just a project, this endeavor epitomizes my relentless pursuit of knowledge and my unwavering commitment to excellence. It is a testament to the transformative power of perseverance and the boundless possibilities that await those who dare to dream.