The Silent Backbone Why Traditional Machine Learning Still Matters in the AI Era
Machine Learning AI Data Science
A deep dive into the relevance of traditional machine learning techniques in the age of advanced AI.
4/22/2025
Read More →
Yuval Mehta
AI/ML Engineer | Deep Learning Practitioner | MLOps Enthusiast
I specialize in building scalable machine learning pipelines, computer vision models, and NLP systems that deliver real-world impact. Currently working on intelligent systems at the intersection of AI research and production-ready deployment.
About Me
Hi, I'm Yuval Mehta 👋
I'm a passionate AI/ML Engineer driven by curiosity and impact. I specialize in developing data-driven solutions that blend machine learning, deep learning, and scalable systems to tackle real-world challenges. With hands-on experience in computer vision, natural language processing, and MLOps, I bring innovative AI models from research to deployment.
🚀 What I Do
- ML System Design – Building and scaling end-to-end ML pipelines using MLOps best practices
- AI Development – Creating production-ready models in computer vision and NLP domains
- Experimentation & Optimization – Fine-tuning models for peak performance and reliability
- Product Impact – Developing intelligent systems that deliver measurable business outcomes
🏆 Key Achievements
- Ranked Top 1% (274th/74,824) in Amazon ML Challenge 2024 with a team of four
- Published research paper on Early Diabetes Prediction at IEEE APCIT 2024
- Improved video processing throughput by 30% at IIT Kharagpur using GNNs and autoencoders
📚 Education
- B.Tech in Computer Engineering
- NMIMS - Mukesh Patel School of Technology, Management and Engineering
- Specialization in Artificial Intelligence
- Focused on: Machine Learning, Deep Learning, Computer Vision, Natural Language Processing, Reinforcement Learning and MLOps
- Core: Data Structures and Algorithms, Computer Architecture and Organization, Operating Systems, Database Management Systems, Computer Networks, Theory of Computation
💻 Technical Skills
Languages:
Python, JavaScript, SQL, C, C++, Java
ML/DL Frameworks:
PyTorch, TensorFlow, Keras, Scikit-learn, XGBoost, OpenCV, Transformers, NLTK, spaCy, LightGBM, CatBoost
AI Agents:
LangChain, Langgraph, CrewAI, LLamaIndex
MLOps & Cloud:
MLflow, Docker, Weights & Biases, DVC, CI/CD, ONNX, AWS, GCP, Azure
Web Technologies:
Django, FastAPI, Flask, Streamlit, Node.js, Express.js, HTML, CSS, Streamlit
Databases:
MySQL, MongoDB, PostgreSQL, Supabase, Vector Databases, Redis, Pinecone, ChromaDB, FAISS, SQLite
Automations:
N8N, Make, LangFlow, Zapier
Big Data & Distributed Computing:
Pyspark, Apache Spark
Specialties:
Computer Vision, NLP, Generative AI, ML Pipelines, Reinforcement Learning, LLM Fine-tuning, AI Automation, Agent-based Systems
Experience
AI/ML Intern
xLM - Continuous Intelligence - Mumbai, Maharashtra
January 2025 - May 2025
- Optimizing AI-driven automation and machine learning models for improved performance.
Machine Learning Intern
IIT Kharagpur (Remote) - Kharagpur, West Bengal
July 2024 - May 2025
- Conducted experiments with auto-encoders and GNN, boosting video processing by 30%.
- Optimized hyper-parameters, resulting in a 20% increase in model accuracy and efficiency.
Machine Learning / Data Science Intern
JM Financial Ltd - Mumbai, Maharashtra
July 2024 - November 2024
- Automated KYC processes using computer vision and deep learning, cutting processing time by 40%.
- Developed OCR solutions to enhance document verification efficiency by 30%.
- Analyzed large data to derive actionable insights, influencing strategic decisions and operational efficiency by 15%.
Backend Developer Intern
Kenmark ITAN Solutions - Mumbai, Maharashtra
December 2022 - April 2023
- Engineered robust APIs that boosted integration efficiency by 30% across multiple platforms.
- Implemented comprehensive QA protocols, enhancing overall system reliability by 20%.
- Revamped SQL and MySQL databases, significantly cutting query response times by 15%.
Featured Projects
AQI and Molecule Prediction
ViT-based model to predict AQI and molecular data from satellite and street view imagery. Optimized for accuracy and cloud scalability.
PyTorch MLFlow GCP ViT Computer Vision Multi-modal
Verbal-Vision
Lip reading model with 87% character-level accuracy. Deployed using TorchServe with optimizations for inference and training time.
PyTorch MLflow TorchServe Computer Vision NLP MLOps Docker
Image-Lingo
Image captioning model using attention mechanism with 91% accuracy. Integrated MLOps for model tracking and performance improvements.
PyTorch AWS MLflow Attention
Outreach-Ace
Gen-AI based resume analyzer and cold email generator with Langchain and LLMs. Includes a user-friendly Streamlit interface.
Langchain LLM ChromaDB Streamlit Gen-AI
Urban-Echo
Sound classification model trained on 8,000+ audio samples. Enhanced model accuracy and training using audio preprocessing and MLflow.
Audio PyTorch MLflow Docker
Latest Blog Posts
The Rise of Intelligent Enterprises AI at the Heart of Business Strategy
AI Business Strategy Innovation
An exploration of how AI is becoming a core component of business strategy, driving innovation and efficiency.
4/11/2025
Read More →
AI-Powered Digital Twins The Future of Intelligent Systems and Real-World Optimization
AI Digital Twins Optimization
A comprehensive look at how AI-powered digital twins are transforming industries by enabling real-time monitoring and optimization.
4/8/2025
Read More →
Scaling ML Models- The Hidden Challenges No One Warned Me About
Machine Learning Scalability MLOps
An exploration of the often-overlooked hurdles in scaling machine learning models effectively.
3/24/2025
Read More →
Smarter Diagnostics- How AI is Enhancing Accuracy and Speed in Healthcare
AI Healthcare Diagnostics
A look into how artificial intelligence is revolutionizing diagnostic procedures in the healthcare industry.
2/16/2025
Read More →
Demystifying Transformer Models- A Deep Dive into NLP’s Revolutionary Architecture and Its Real-world Applications
NLP Transformers Deep Learning
Explore the evolution, components, and real-world applications of transformer models in NLP and beyond.
1/26/2025
Read More →
Research & Publications
Examining ML Approaches for Early Diabetes Prediction
Authors: Yuval Mehta, Team
Conference: IEEE Asia Pacific Conference on Information Technology (APCIT) 2024
Abstract: This research explores multiple machine learning models for early diabetes prediction, highlighting key patterns in patient health data to aid in proactive healthcare measures.
View Publication →Contact Me
I'm always open to new opportunities and collaborations. Feel free to reach out!