About me

Machine Learning & MLOps Engineer with 8 years of experience in building and deploying advanced AI solutions across cloud environments. My expertise includes designing, optimizing, and automating scalable ML pipelines using Python, Docker, and Kubernetes.

I specialize in NLP, Computer Vision, and RAG systems , with hands-on experience in TensorFlow, PyTorch, and AWS SageMaker. My focus extends to cloud solutions on AWS, GCP, and Microsoft Azure, ensuring high-performance, cost-effective, and resilient architectures.

With expertise in CI/CD automation , model lifecycle management, and cloud infrastructure, I ensure seamless deployment and monitoring of machine learning models. As an AWS Certified Solutions Architect - Professional , I design and implement efficient cloud architectures, ensuring scalable and secure machine learning deployments.

Experience

• MLOps Engineer - AI-Driven Job Recommendation & Resume Parsing System

  RapidScale — 04/2023 - 04/2025

  Developed and deployed scalable ML pipelines using AWS SageMaker, Docker, and Kubernetes for job recommendation and resume parsing. Built optimized data pipelines using AWS Lambda for real-time data collection and AWS Glue for transformation. Created embedding models using Sentence-BERT and AWS SageMaker for semantic similarity matching. Implemented CI/CD workflows with Jenkins and Terraform for automated deployment and versioning. Optimized inference pipelines using TensorFlow Serving and Docker containers. Stored embeddings in AWS S3 and DynamoDB for efficient retrieval. Monitored model performance with AWS CloudWatch and SageMaker Model Monitor. Built RESTful APIs using FastAPI to serve embedding-based job recommendations.

  
  • AWS SageMaker
  • Docker
  • Kubernetes
  • Python
  • TensorFlow
  • FastAPI
  • AWS Lambda
  • AWS Glue
  • Sentence-BERT
  • Jenkins
  • Terraform
  • AWS S3
  • DynamoDB
  • AWS CloudWatch
• ML Engineer - AI-Powered Conversational Assistant

  TikTok — 04/2022 - 02/2023

  Spearheaded development of an AI-powered conversational assistant supporting creators and enhancing in-app user experience across millions of global users. Customized and fine-tuned Large Language Models (LLMs) using TikTok-relevant data for multilingual understanding. Designed smart retrieval system (RAG) blending vector similarity search with graph-based user interaction data (Neo4j), improving contextual relevance by 30%. Built tools understanding slang, emojis, and cultural nuances using cutting-edge NLP and NLU techniques. Collaborated with content safety teams to implement AI-assisted moderation flows. Leveraged Azure AI and Databricks for large-scale experimentation and deployment. Integrated Azure OpenAI services for generative capabilities. Achieved 20% latency reduction for LLM inference through GPU acceleration optimization.

  
  • LLM
  • RAG
  • Neo4j
  • NLP
  • Azure AI
  • Azure Databricks
  • Azure OpenAI
  • Python
  • PyTorch
• Data Scientist - Credit Risk & Fraud Detection

  Danske Bank — 01/2021 - 03/2022

  Developed machine learning models for credit risk assessment, fraud detection, and customer segmentation using Random Forest, XGBoost, and Neural Networks. Preprocessed and transformed data using advanced feature engineering, missing data imputation, and normalization. Implemented time-series forecasting models for financial market predictions using ARIMA and LSTM networks. Optimized model performance through hyperparameter tuning, cross-validation, and grid search. Automated end-to-end data pipelines using Python and Apache Spark, significantly reducing preprocessing time. Collaborated with business analysts and financial teams, ensuring models aligned with business objectives and compliance regulations. Presented findings using Tableau and Matplotlib for visualization.

  
  • Python
  • Random Forest
  • XGBoost
  • Neural Networks
  • ARIMA
  • LSTM
  • Apache Spark
  • Tableau
  • Scikit-Learn
• Computer Vision Engineer - CoreML Mobile Solutions

  SenseTime — 09/2019 - 12/2020

  Developed and optimized CoreML-based computer vision models for real-time inference on iOS devices, powering visual intelligence features in consumer-facing mobile apps. Contributed to SenseNova foundation model system by integrating lightweight yet high-performance visual models into mobile environments. Leveraged SenseCore AI infrastructure to streamline model training and deployment pipelines. Engineered and converted deep learning models (CNNs, object detection, segmentation) into CoreML format with quantization, pruning, and compression. Collaborated with iOS engineers to embed real-time vision capabilities including face tracking, scene recognition, and AR overlays. Designed custom preprocessing and postprocessing logic, improving prediction accuracy and reducing latency by over 25% across iPhone and iPad devices. Enabled use cases across Smart Auto, AI-powered retail, and mobile AR.

  
  • CoreML
  • Computer Vision
  • Deep Learning
  • CNNs
  • iOS
  • Object Detection
  • Segmentation
  • OpenCV
  • C/C++