Zaid Abidi

Data Scientist

India2 yrs 11 mos experience

Most Likely To SwitchAI ML Practitioner

Key Highlights

Expert in generative AI and computer vision.
Developed models achieving over 94% accuracy.
Proven track record in automating complex processes.

Stackforce AI infers this person is a Data Scientist specializing in AI and Machine Learning for Fintech and SaaS applications.

Contact

Skills

Core Skills

Large Language Models (llm)Machine LearningComputer Vision

Other Skills

TransformersWord2VecPytesseractBERTLayoutLM V1JSONRSSDocument RepresentationClustering techniquespandasNumPyPySparkScikit-learnPyTorchTensorFlow

About

I am a full-stack Data Scientist specializing in generative AI and computer vision who is characterized by his versatility, adaptability and passion for innovating new out-of-the-box methods to solve real-world problems utilising the power of cutting-edge technologies like Artificial intelligence, cloud computing and big data analysis. My desire for problem-solving is only eclipsed by my neverending thirst for knowledge on various topics such as the latest technological breakthroughs in tech. I have three years of experience in using top-of-the-line computer vision models like vision transformers and CNN models like EfficientNetV2 to develop systems for pose detection, recommendation, object detection, etc. I am also well-versed in leveraging recent advancements in generative AI to develop LLM-based chatbots using BERT and Langchain, I am also well well-versed with text-to-image models like stable diffusion.

Experience

2 yrs 11 mos

Total Experience

1 yr 5 mos

Average Tenure

2 yrs 1 mo

Current Experience

Zs

Advanced Data Science Associate

Apr 2024 – Present · 2 yrs 1 mo · Bengaluru, Karnataka, India · Hybrid

Large Language Models (LLM)

Intellect design arena ltd

Data Scientist

Jun 2023 – Apr 2024 · 10 mos · Chennai, Tamil Nadu, India · On-site

Intelligent Document Processing System
● Aim: Elimination of human involvement in the handling and classification of substantial volumes of confidential documents.
● Technology Used: Transformers, Word2Vec, Pytesseract, BERT, LayoutLM V1, JSON, RSS, Document Representation, Clustering techniques.
● Results: Developed a production-ready model for an intelligent document processing system that automates document processing, reducing personnel expenditures by 100% and increasing time saved by more than 5000%.
Customer Churn Prediction for Wealth Management
● Aim: Develop a machine learning model to predict customer churn in wealth management, enabling proactive retention strategies.
Technology Used:
● Data Processing( pandas, NumPy), Data wrangling, cleaning, massaging techniques, PySpark (for large-scale data processing), Machine Learning,(Scikit-learn: logistic regression, SVM, random forest), feature importance, PyTorch, TensorFlow (for deep learning models like ANN, CNN, LSTM, BiLSTM, CNN+LSTM), XGBoost, Feature Engineering, Model Selection, Hyperparameter Tuning, Grid Search, Randomized search, AutoML Libraries (Optuna, Hyperopt, Raytune), Docker, Kubernetes, ROC AUC, Chi-square tests, correlation checks, AWS SageMaker, AWS S3, AWS ECR, Data visualization (Matplotlib, Seaborn)
● Results: Developed a model with 94.63% accuracy in identifying churn risk. This is expected to lead to an 80% reduction in customer churn, resulting in millions saved for the wealth management firm.