Sandip P.

Hello, I'm Sandip Patel, a passionate and experienced data scientist with expertise in predictive modeling, statistical modeling, and clustering models. I possess a strong background in developing both supervised and unsupervised machine learning models, with a particular interest in deep learning models. My journey in the field of AI has led me to excel in various domains, including chatbot creation, text summarization, NLP tasks, and text classification. I take great pride in my ability to tackle complex challenges and ensure high-quality data by applying my data cleaning skills. Throughout my career, I have honed my proficiency in a wide array of technical libraries and frameworks, including Keras, Pandas, Numpy, TensorFlow, Scikit Learn, Scipy, spacy, gensim, nltk, matplotlib, and seaborn. These tools have been instrumental in achieving remarkable results and driving insights from data. If you are looking for a data scientist with a strong foundation in machine learning, an eye for detail in data cleaning, and a knack for developing efficient models, I would be thrilled to collaborate with you on your next project. Expertise: ✅ Predictive modeling, ✅ Statistical Modeling, ✅ Clustering models, ✅ Supervised/ Unsupervised Models, ✅ Deep Learning models, ✅ Chatbot creation, ✅ Text Summarization, ✅ NLP Tasks, ✅ Text Classification, ✅ Data Cleaning, ✅ ML Modelling, Technical Libraries and Frameworks: ✅ Keras ✅ Pandas ✅ Numpy ✅ Tensorflow ✅ Scikit Learn ✅ Scipy ✅ spacy ✅ gensim ✅ nltk ✅ matplotlib ✅ seaborn Let's work together to unlock the full potential of your data !! PROJECTS: 1. AI Interview in Recruitment Platform ▣ The application is to build AI based requirement platform where AI helps to match candidates profile with the job post. ▣ Candidate can be evaluated by AI interview and shortlisted. ▣ There will LLM integration behind which generate the question and ask to candidate. AI also evaluate the answer of the candidate and then give summary. Skills : Python, Prompt Engineering, DeepGram, Fast API, React, AWS 2. TrichoAI - Hair loss Classification and Detection ▣ Developed a machine learning model which classified hair diseases from different 9 categories and also classify the stage as ‘Erly Stage’, ‘Late Stage’, ‘No Hair Loss’. ▣ Machine learning model also detect the area of diseases as well. ChatGPT API integration to prepare statement which describe disease in nicely manner from the model output. ▣ The accuracy of the model was 85-87% and right not deployed in the production site with mentioned url. Skills : Computer Vision, Deep Learning, Machine Learning Model, Python, FastAPI 3. AirVideo - AI Based Youtube Comment Analyser ▣ AirVideo is a platform designed for analyzing YouTube comments, where YouTubers or other users can register to examine the comments on their YouTube videos. ▣ The platform offers various analytical tools, including sentiment analysis, identification of positive and negative comments, discovery of new ideas, detection of humor tones, content suggestions, summaries of comments, and a list of the top 10 most popular comments. ▣ The application utilizes its own customized large language model (LLM) to analyze the text of YouTube comments. Skills : Python, NLP Tokenization, LLM Prompt Engineering, Node.js, OpenAI API, Stripe Payment, AWS 4. Doctor-Patient Speech to Text Transcriber Application ▣ The project was to develop a doctor-patient collaboration application. In this application, doctors could provide prescriptions and other details verbally, and the system would convert these from speech to text using the Deepgram API. ▣ Due to data privacy concerns, we decided against using any third-party APIs and instead opted to create our own model for transcribing and translating texts. The application's role was to identify health-related terms and generate reports. ▣ The system was generating the recordings in mp3 file to further analysis. Skills : LLM Prompt Engineering, OpenAI API, NLP Tokenization, FastAPI ,AI Audio Generator, Azure 5. Telebot an AI Agent (LLM, LongChain, RAG based solution) ▣ Telebot, an AI agent based on LLM, LongChain, and RAG technologies. ▣ The purpose of Telebot was to initiate calls and manage conversations with clients in the Real Estate sector. The system converted the client's spoken words into text, then used a customized LLM model to generate responses, which were subsequently converted back into speech. ▣ The system could generate responses from previously saved conversations. The entire process is taking approximately 3-4 seconds. The system featured speech-to-text and text-to-speech capabilities. Skills : NLP Tokenization, LLM Prompt Engineering, AI Agent Development, Text Analytics, Machine Learning. NOTE : YOU CAN FIND ALL PROJECTS IN THE PORTFOLIO