• Experienced AI Engineer and Data Scientist with 5+ years of experience.
• Extensive experience in Python programming language, utilising it for various AI and data-related projects.
• Worked on Object Detection techniques, leveraging AI models to identify and locate objects in images and videos. I have good experience with OpenCV, NLTK and spaCy.
• Proficient in Keras, having used it to build and train deep learning models.
• Worked on data manipulation and analysis using NumPy and Pandas, ensuring effective data processing.
• I have a good understanding of Amazon RDS, SQL warehouse and familiarity with MongoDB.
• Experienced in Exploratory Data Analysis (EDA) and Data Visualization, gaining insights and communicating data findings effectively.
• Worked on Generative Adversarial Networks (GAN) and Generative Art using AI, generating new data instances and creating artistic creations.
• Experienced in LSTM, RNN, and NLP, utilising them for sequence modelling, time-series predictions, and natural language processing tasks.
• Developed Chatbot applications, leveraging AI technologies to create interactive conversational agents.
• Proficient in TensorFlow, using it for building and deploying machine learning models.
• Worked on creating APIs and deploying models for seamless communication between software systems and real-world ML applications.
• Experienced in Optical Character Recognition (OCR) and Computer Vision, extracting text from images and analysing visual information.
• Worked on MLOps and Docker, managing the lifecycle of machine learning models and streamlining application deployment.
• Exposure to Pyspark and other ETL tools.
• Utilised AWS and Cloud Computing for scalable and reliable cloud-based solutions.
• Basic knowledge and experience in Azure DataBricks, Dlib, PocketSphinx, HuggingFace and Textblob.
• Developed Flask APIs, creating web applications and APIs using the Flask framework in Python.
• Proficient in MySQL, ensuring efficient data storage and management in various applications.

Python
Object Detection
OpenCV
NLTK
Keras
NumPy
Pandas
EDA
spaCy
Apache
Kaa
Data Visualization
GAN
Generative Art using AI
LSTM
Chatbot Development
TensorFlow
RNN
API
Model Deployment
OCR
Computer vision
NLP
MLOps
Docker
AWS
Cloud Computing
Flask API
MySQL

Genie Bot
Tools & Technologies: Python, Data Extraction, LSTM, NLP, Data Cleaning, Jupyter, Apache Spark, NLTK, Tensorflow, Scikit-learn.
Description: Build an automation script for an in-house Conversational Chatbot base model, which reduced production time by 80%. Data cleaning after creating queries for our project developed an end-to-end conversational chatbot using a base model for 3+ international clients, implementing advanced NLP techniques consisting of domain and intent classification and entity
recognition. LSTM and BiLSTM models are used for entity detection. Build a prototype model of a conversational and rule-based chatbot to meet the client's requirements. Created a POC of the base model chatbot on MLOps platforms Airflow and MLFlow.

Human-to-anime for Metaverse product
Tools & Technologies: GAN, CNN, Transform Data, Object detection, Data Gathering, OpenCV, Extract Data, PyTorch, Keras, Scikit-learn, Matplotlib.
Description: We created human-to-anime images for the Metaverse product, in which we used the unsupervised image-to-image translation technique model UGATIT. Achieved 75% accuracy through AdaIN. Extract human and anime images from different open-source platforms. Clean raw images and filter data from the database.

Artificial NFTs for XANALIA Product
Tools & Technologies: Python, Data Cleaning, Pandas, Data Gathering, CNN, OpenCV, YOLO, OpenCV.
Description: We built a CNN-based model for a decentralised NFT marketplace using the neural-style TF technique and generated 10k+ bizarre NFTs for the XANALIA Platform, whose market cap is more than 10 million USD.

Multiple POC project
Tools & Technologies: Python, GAN, Data Cleaning, Pandas, Data Gathering, CNN, OpenCV, Data Pipeline, Tensorflow, SQL, ETL, Python, PyTorch.
Description: Improve low-quality images into high-quality using SR-GAN and basic OpenCV functions. We achieved 50% accuracy in the virtual try-on model "Down to the Latest Detail", It was possible only because of these 3 things. Pose keypoints, semantic parsing, and a cloth mask. The SSIM evaluation is 0.71. Contribution to the Unity team for 3D mesh, experimental practises on different models like pifuHD and pytorch3DIt reduced the time of mesh generation by 30%. Delivered 20+ POCs and their technical documentation with the CTO and team lead. Ideated a new approach for product automation and
visualisation using ML/DL, reducing time by 70%. I created a script for the image classification model for data cleaning, and after this, I passed the dataset.