Personal details
Vineel K.
Summary
Data Scientist|Software Developer|5 years experience|NLP|AWS|GCP|Kubernetes GPU
• Working on end-to-end Machine Learning life cycle, data pipelines and distributed computing platform. Developed ML platform for model training and deployment in production.
Currently processing around 10 million inference requests per day. Developed an open source python package library(onepiecepredictor) and NLP AI-based chatbot using Tensorflow, Sci-Kit Learn, google dialogflow PHP, python flask, neo4j for knowledge graph.
Productionised a chit-chat bot using deep learning sequence to sequence model with LSTMS, attention using TensorFlow GPU. Helped clients make strategic marketing channel investments by building interactive and automated KPI reports with data from various sources using MySQL.
Developed big data pipeline to transfer large volumes of Ad auction requests in S3 to Amazon Redshift cluster using spark. Has a solid software development and data analysis experience of 3 years using Microsoft .NET Web API, C#, Microsoft SQL Server, REDIS, SignalR and RabbitMQ and JavaScript.
- Skills: C++, C#, Python, data structures and algorithms, R, MATLAB, HTML, JavaScript, React JS, SQL, NoSQL, Hadoop, Apache Spark, Neo4j, Machine Learning Algorithms, deep learning, statistics, data modelling, predictive modelling, Object-oriented design patterns, Unit testing
Technical skills
Work Experience
• Working on end-to-end Machine learning models lifecycle, data pipelines and MLOps. Developed ML platform for model training and deployment in production. Currently processing around 20 million inference requests per day.
• Created Machine Learning infrastructure and inference as service in production environment using Airflow, Dockers, Kubernetes GPU cluster on GCP with auto-scaling. Developed python flask REST APIs, CI/CD pipelines ETL data pipelines for Computer Vision deep learning model’s inference to clean and order vehicle images in auto-motive dealer ship websites.
• Developed APIs, micro-services for API rate limiting, authentication, request logging using Python, MySQL, REDIS, BigQuery, Tensorflow, PyTorch ,OpenCV, sklearn, Pandas. Created billing, performance, usage analytics data pipelines and dashboards.
• Helped vehicle dealerships websites build statistical AB Testing pipeline for testing click rates with different backgrounds.
• Trained and deployed NLP AI Chatbot in production on AWS cloud. Performed text pre-processing and EDA. Trained deep learning LSTM, BERT Transformer and text summarization models. Used Python, Keras, NLTK, Huggingface, Google Dialogflow.
• Developed python big data SQL ETL Apache spark pipeline on AWS using Apache Airflow to transfer large volumes of Advertisement auction requests in AWS S3 buckets to Amazon Redshift cluster.
• Helped vehicle dealerships make strategic marketing investments by building automated KPI dashboards with data from various sources using Python, MySQL, data pre-processing, cleaning techniques. Reduced clients spending by around 20%
- Implemented real time event streaming and messaging services using RabbitMQ, SignalR, .NET, C# to help push cache invalidation events, notifications and admin messages across application.
• Improved application scalability and performance by tuning SQL server queries and implementing distributed parallel computing services using C#. Implemented code using modern Object-Oriented Design principles.
• Implemented a real time dependency identification using Neo4j graph database for optimised and incremental data process.
• Built Data Analytics OLAP cube using SSAS on HR/Workforce financial data. Increased the new user acquisition rate by 20%.
• Developed Web Service REST API's to manipulate employee’s data in bulk, which runs 50% faster than existing ETL pipeline. • Developed interactive KPI’s dashboards for quick data insights. Used React JS, .NET web API, C#, REDIS, SQL server.