Name: Ayan Maity
Current Role: Software Engineer
Email Id: ayanmaity813@gmail.com
Address: Bangalore, India
Skills
Programming Languages: Python
ML/DL Frameworks: PyTorch, TensorFlow, Keras, Scikit-learn, Transformers (Hugging Face)
LLM & GenAI: LangChain, LangGraph, RAG
Data Science Tools: NumPy, Pandas, Matplotlib, Seaborn, Jupyter
Databases: SQL, Neo4j, Qdrant
Version Control: Git, GitHub, Bitbucket
Other: R, Excel, PowerPoint, LaTeX
I have completed my B.Sc. in Mathematics from Ramakrishna Mission Residential College, Narendrapur, and pursued my M.Sc. in Big Data Analytics from Ramakrishna Mission Vivekananda Educational and Research Institute, Belur Math. My studies have provided a strong foundation in mathematical principles and statistical modeling, which I have applied in various machine learning, deep learning, and time series forecasting projects.
I completed a Software Developer Internship at Sensight Technologies Pvt. Ltd., Bangalore and am currently working as a Software Engineer at the same company. I possess strong skills in data analysis and manipulation using Python and R, and have hands-on experience with LLMs, GenAI, and machine learning algorithms applied to real-world problems. I am committed to continuous learning and seek to leverage my skills in Data Science, AI, and Data Analysis to contribute to innovative solutions.
My aspiration is to become a leading expert in LLM, Data Science, Machine Learning, Deep Learning, Natural Language Processing and Time Series Forecasting domains.
Below are the projects that I completed during my Master's program.
Detected Monolingual Code Clone using Graph Matching Network. Trying to perform fine-tuning over CodeBERT and then detect Multilingual Code Clone using that.
Developing a multilingual speech-to-text pipeline, starting with Hindi speech-to-text conversion and extending it to translate into English and generate English speech. Goal is to scale the system for other languages to English translation.
Identified chemical compounds with similar structures using Tanimoto coefficient and Graph Edit Distance algorithms. Predicted biological activities of chemicals based on their structure. Estimated toxicity levels of chemicals using Regression models.
Firstly, conducted an Exploratory Data Analysis to visualize key insights. Handled class imbalance using various methods: SMOTE, TomekLinks, SMOTETomek, ADASYN etc. Applied and evaluated multiple machine learning classification models (e.g. Random Forest, SVM etc.) to predict whether an employee will leave his / her job or not.
Cleaned and normalized the dataset for high-quality input. Designed a Transformer-based architecture for the Text-to-speech synthesis task and fine-tuned hyperparameters for better performance. Also used post-processing techniques like mel-spectrogram inversion and waveform synthesis to improve audio quality.
Conducted EDA, time-series indexing, and handled missing values. Identified seasonal patterns and trends with decomposition techniques. Applied ACF and PACF plots, Rolling Statistics, ADF test for checking stationarity. Implemented SARIMA and AutoARIMA models for forecasting, tuned model hyperparameters, and validated forecasts by comparing predicted sales to real sales.
This is an entity-level sentiment analysis dataset of twitter. Given a message and an entity, the task is to judge the sentiment of the message about the entity. This project involves building a sentiment analysis model using a dataset of Twitter posts. The goal is to classify the sentiment of each tweet as positive, negative, neutral or irrelevant.
This project explores the impact of various factors on the placement status of engineering college students. The factors analyzed include internships, projects, workshops, aptitude scores, SSC and HSC marks, extracurricular activities, placement training, and soft skills ratings.
Below are the details to reach out to me!
Kodihalli, Bangalore, Karnataka - 560008, India