Ramu Vootla

With over 15 years of extensive professional expertise spanning various IT domains, I am a dedicated machine learning professional and data-driven analyst. My proficiency lies in effectively applying machine learning techniques and harnessing algorithms to address tangible business challenges. I possess a comprehensive understanding of supervised, unsupervised, and reinforcement learning algorithms, and I am adept at implementing diverse algorithms such as linear regression, decision trees, random forest, support vector machines (SVM), logistic regression,

Additionally, I bring hands-on experience with cloud platforms such as Google Cloud Platform (GCP), and Amazon Web Services (AWS), along with familiarity with OpenAI, Gen AI, Gemini Pro, Gemini pro vision, Hugging Face and FB Large Language Models (LLMs). Motivated and tech-savvy, I have a profound capability to swiftly grasp and integrate new technologies, showcasing a strong commitment to staying at the forefront of the ever-evolving tech landscape.

In my role as an MLOps Engineer, I bring extensive expertise in working with AWS SageMaker, TensorFlow, PyTorch, _google_genai, and employing advanced techniques such as distributed training and transfer learning. Proficient in handling substantial datasets, and . My data analysis and visualization capabilities extend to platforms such as  Apache Spark. Moreover, I excel in MLOps, cloud computing, and employ tools like Terraform, GitHub, Git, Docker construct efficient data pipelines. My proficiency extends to deploying models both on-premises and in the cloud, employing Lambda and Step Functions for seamless integration

MLOPS Engineer  | TATA Consultancy services | September 2020 - Present

• Worked for various clients (USAA,MARS,GSM) as Machine learning  and MLOps engineer.
• Automated AWS infrastructure by creating Terraform modules and registered in Terraform Enterprise Registry  along with versioning.
• Implemented an automated end-to-end model deployment process utilizing Kubernetes clusters, with a focus on creating dedicated pods for effective model     monitoring and logging within a dynamic environment.
• Developed a Python Software Development Kit (SDK) to facilitate token-based authentication for seamless check-in of models into the model repository, ensuring secure access based on predefined roles.
• Possess a robust understanding of AI principles and techniques, with a focus on natural language processing. Demonstrated experience in the development and optimization of Generative AI systems for real-world applications.
• Demonstrated expertise in seamlessly integrating and enabling Language Models (LLMs) within existing cloud infrastructure, highlighting a skill set in optimizing AI solutions for enhanced performance and scalability.
• Ability to troubleshoot issues in deployment environments, clusters, pods, containers, and Docker images, showcasing proficiency in resolving complex technical challenges within the deployment pipeline.
• Having experience in creating sagemaker domain with Identity Center as well as IAM in VPC mode.
• Experience in creating and spinoff instances, buckets, glue jobs and other AWS resources using Terraform.
• Implemented Lifecycle configuration scripts and attached to sagemaker domain and limiting access to the instance types as well as kernels.
• Having hands on experience in programming using TensorFlow, PyTorch, Python Programming.
• Implemented a CI/CD pipeline using GitLab to automate the model deployment process and streamline the development cycle.
• Utilized Python programming to build custom scripts for deployment and automation.
• Involved in end-to-end implementations for various ML projects like AR Forecasting and Amazon Lex models.
• Applied various transfer learning techniques using existing pre-trained models like EfficientNetB0 for image classification, word embeddings like Glove, BERT, Universal Sentence Encoder for unstructured text for NLP. Developed NLP models for Topic Extraction and Sentiment Analysis. Developed internal text processing libraries using nltk and spacy.
• Large amount of information retrieval, extraction, translation, text simplification and summarization (removing punctuation, converting text to lowercase, removing stop words) and preprocessing using stemming and lemmatizing techniques.
• Vectorizing data with bag of words, N-Grams, TF-IDF techniques and finding patterns using various embedding and tokenization techniques using NLP and nltk.BertEncoder,Embeddings, Padding mask, SpaCy, NLTK libraries.
• Proficient in utilizing Generative AI techniques and frameworks such as GANs to generate realistic and creative content with a deep understanding of advanced techniques including GPT (Generative Pre-trained Transformer), T5 (Text-To-Text Transfer Transformer), NLP techniques to preprocess, analyze, and transform textual data, enabling effective input to generative AI models.
• Demonstrated experience in working with open AI technologies. and leveraging their capabilities to develop innovative solutions.
• Good knowledge of recurrent neural networks, LSTM networks and word2vec. - Targeting multiple models to get trained in the same EMR clusters to reduce cost.
• Created customized docker images and pushed to ECR and populated  in jupyter notebooks - Handling large amount of data using Vaex data frames and Parquet files.
• Worked on data analysis, analysis reports and ETL operations using powerbi, tableau splunk and glue connectors.
• Implemented distributed training with TensorFlow and sagemaker by GPU instances.
• Using AWS Athena for writing range queries and Glue for S3 data partition.
• Provisioning AWS resources and maintaining state using Terraform commands. - Downloading multiple files from S3 in parallel using multiprocessing and multithreading
• Hosted Flask and FastAPIs in dockers and on premises IIS.
• Trained models in Persistent and Transient EMR GPU clusters.
• Executing R Scripts from python in conda and python environments.
• Using JupyterHub, Zeppelin notebooks, Spyder IDEs for implementing ML models. - Using BigDataLake portal and boto3 SDK to upload files into S3 bucket.
• Extensively used Apache Spark, Hadoop, pyspark, Hive and presto.
• Played lead MLOps role for End-to-End MLProcess. Responsible for deploying models in on premises and AWS cloud.
• Giving complete End to End demos on sagemaker, GitHub, git, APIs, OpenAPI specification, CICD pipeline to Data Science team.
• Implemented data pipelines and DAGS using Apache airflow.
• Implemented Amazon API gateways, triggers deployed lambda functions which internally calls deployed models for inference.
• Implementing fast APIs and calling the models and getting the predictions and return them. - Connecting to BigDataLake using pyodbc in Zeppelin notebooks and extracting data into spark dataframe, transferring data from spark dataframe to parquet files. Saving parquet files into S3 bucket using boto3 SDK. Downloading parquet files from S3 bucket and converting them into TensorFlow datasets for further processing.
• Deployed models with sagemaker pipelines and invoked these pipelines with terraform scripts.

• TCS Associate Consultant June-2010 to Till Date
• DHII Health Tech Technical Lead Sep 2009 - May 2010
• IDOLAFORI Design Pvt. Lts .NET Developer Feb 2008 - June 2009
• Cible Technologies .NET Developer Nov 2006 - Feb 2008

• Automation Anywhere - Advanced RPA Professional Course A2019
• Uipath Certificate – on/Advance Training
• Microsoft .NET Framework 3.5 ASP.NET, WCF applications
• Microsoft .NET Framework 2.0 - Application Development Foundation Certification
• MCTS: .NET Framework 4, Service Communications Applications Certification