Hi, I am Sneh Shah.
Aspiring Data Scientist
I am a highly motivated ML Engineer at Salesken.ai with a strong background in Data Science. With a competitive nature, I passionately solve problems and build AI solutions 🤖, showcasing excellent problem-solving skills and a proven ability to adapt quickly to new technologies 🚀.
Mail: snehshah2901@gmail.com
Contact: +91 9428911398
Experience
ML Engineer 1
Salesken.ai, Bengaluru, Karnataka
Apr 2024–Present
- Engineered an Email Generation System using LLM Agents with custom prompts and personalized responses for clients, improving custom email creation accuracy.
- Designed a Knowledge Extractor System using RAG architecture, integrating web scraping, Azure services, and Kafka, enhancing data retrieval efficiency by 25%.
- Introduced a PII masking solution utilizing pseudonymization, anonymization, tokenization, and truncation techniques, ensuring compliance with data privacy regulations.
- Constructed a DS Toolkit providing API endpoints for call duration analysis and deal/task summarization using LLM, increasing efficiency by 40%.
- Designed a Kafka Controller for efficient message handling, idempotency, reducing data duplication by 15%.
- Optimized the Whisper-ASR system and deployed it on the Triton Inference Server, reducing latency by 25% and enhancing scalability.
Technologies:
- Python
- LLM
- Langchain
- Azure
- Kafka
- Triton Inference Server
- Kubernetes
- VectorDB
- MongoDB
- Postgres
- Streamlit
Projects
Neural Style Transfer
Orchestrated the research and development of a Neural Style Transfer project, analyzing recent advances and 2 main techniques. Engineered a high-performance NST system with approximately 30% reduction in processing time using L-BFGS optimization.
- Python
- Pytorch
- Deep Learning
- Artistic Image Enhancement
- L-BFGS optimization
Diabetic-Retinopathy Detection
This project aims to use Diabetic retinopathy detection using features from deep learning and fitting into machine learning algorithms and the data is taken form the kaggle.
- Python
- Tensorflow
- OpenCV
- Scikit-Learn
- Pandas
CreditCard Fraud Detection
This project is focused on building a credit card fraud detection system using a dataset from Kaggle. The goal is to develop a machine learning model that can accurately detect fraudulent credit card transactions and minimize false positives.
- Python
- Keras
- Scikit-Learn
- Machine Learning
- Matplotlib
Internships & Certifications
Saleken (ML Intern)
• Optimized the Whisper-ASR codebase through Docker deployment, documenting streamlined steps. • Elevated code quality by introducing 2 specialized modules for speaker diarization and voice activity detection. • Successfully deployed Emotion Detection models on Triton Inference Server, optimizing configurations and integrating metrics with Prometheus for effective monitoring.
- Python
- Docker
- Triton Inference Server
- LLMs
- Fastapi
- Hugging Face
- VAD
Open Weaver
• Developed a customized interface for Chat-GPT, acquiring hands-on experience in seamlessly integrating AI language models.• Utilized Open AI's Whisper and DALL-E tools to convert audio into visually generated images.• Applied deep learning algorithms and computer vision methods to detect and mark Deep-Fake images.
- Python
- Gen -AI
- NLP
- Deep Learning
- gradio
- NLTK
- Spacy
Publications
Education
Christ (Deemed to be University)
Masters of Science (Data Science)
Current GPA: 3.63/4
Aug 2022 - Present
Bengaluru, Karnataka
Gujarat University
Bachelors of Science (Data Science)
GPA: 7.77/10
Jun 2019 - May 2022
Ahmedabad, Gujarat
Skills
Languages
- Python
- SQL
- R
- Matlab
- Java
Tools and Frameworks
- Git
- Tableau
- PowerBI
- Weka
- MS Excel
- AWS
Libraries
- Scikit-Learn
- NumPy
- Pandas
- OpenCV
- TensorFlow
- Keras
- Pytorch
- Matplotlib
- Seaborn
- Transformers
- Spacy
- NLTK
Certifications
- Machine Learning Specialization (Jun 2023)
- Data Science with Python (Oct 2022)
- Data Analysis with Python (Jul 2022)