hitesh.nimsarkar.8319 User

AI Architect, Machine Learning Developer, Data Scientist, UI Architect, and Full-Stack Developer with over 6+ years in AI/ML, 8+ years in UI development, currently working at UBS Pune. Proven track record of developing scalable ML systems, forecasting models, and financial AI solutions. Expert in cloud-native deployments, MLOps, and end-to-end product development. Experienced in leading crossfunctional teams and mentoring developers.Core CompetenciesMachine Learning & AI:AI Agent Deployment: Developed and deployed agents in real-world environments, increasing task completion accuracy by ~22% across dynamic scenarios.TensorFlow Inference Optimization: Reduced model inference latency by ~45% while maintaining prediction accuracy, enabling decision-making in production systems.Distributed AI Workload Benchmarking: Analyzed single- and multi-node compute performance; achieved ~30% throughput improvement on distributed training/inference pipelines.LLM Customization with LoRA, PEFT, and Fine-Tuning: Fine-tuned large language models for domain-specific tasks, resulting in ~12–25% increase accuracy in responses.Production-Grade Model Integration: Built end-to-end MLOps pipelines with model versioning and low-latency serving; reduced deployment cycle by ~40%.TensorFlow, Keras, PyTorch, Scikit-learn, BERT, GPT, T5, MLflow, Transformers, LSTMs, SVD, ALS, Recommendation Engines, LangChain,NLP, XGBoost, Time Series, PCA, NER, Word2Vec, Text,SentenceTransformers, Isolation Forest, One-Class SVM, LOF, spaCy, BERTopic, LangChain, Azure GPT-4, PySpark, Flask, Databricks, MLflow,PostgreSQL, Summarization, Pinecone, RAG system.Big Data & Processing:PySpark, Apache Spark, Pandas, NumPy, Dask, Databricks, VectorAssembler, Feature Engineering.UI & Java Enterprise Solutions:Full-Stack Application Development: Delivered end-to-end enterprise applications using React, Java Spring Boot, and Azure 13+years of experienceAI Architect| Machine Learning Developer | UI | Fullstackservices, improving feature delivery velocity by ~35%.Microfrontend Architecture: Implemented modular React microfrontends with independent deployment pipelines, reducing crossteam release dependencies by ~40%.UI Component Standardization: Established reusable component libraries, cutting UI development time for new modules by ~25%.Cloud Integration: Designed and deployed Azure-hosted APIs and databases, enabling scalable and secure multi-tenant architectures.Performance & Reliability Optimization: Reduced API response times and UI load latency by ~30% through caching, query tuning, andfront-end optimization.DevOps & Cloud:Azure (ADF, Blob, Functions, DevOps), Docker, Kubernetes, GitLab CI/CD, Azure Static Web Apps, Nginx, Istio, MLflow, Flask APIs.Frontend & Full Stack:React.js, TypeScript, Redux, RxJS, Node.js, Express, MongoDB, Webpack, HTML5, CSS3, Bootstrap, WebSockets.Backend:Java, Spring Boot, RESTful APIs, PostgreSQL, Azure SQL.