Jul 2021 – Jul 2023 · 2 yrs · Hyderabad, Telangana, India

• Working with one of the most data centric manufacturing processes in the world, helping process engineers make better decisions by developing ML solutions to predict the critical dimensions, find important features to control these critical dimensions.
• Developing image analytics solutions to assist the Real time Defect Analysis teams.
• Guiding the team members towards sustainable and scalable practices of ML solution development.
• Specifics of the projects inlcude:
• Working with composite metrology and device data.
• Model Dry etch process with the aim to predict metrology values.
• Super resolution with wafer image data to aid the metrology operations and cut costs in terms of time and tool usage. #imagesuperresolution
• WIS image classification: Predict if the TEOS deposition process was successful using pre and post WIS images. We used a difference of the RGB images and trained a model to predict the good and bad wafers. Achieved 85.7% precision and 100% recall.
• Defect classification using images captured on manufacturing lines.
• Research and implementation of experiment tracking, model versioning and drift capture mechanisms (MLOps) #mlflow #mlops
• Lead the adoption of code standards and code review mechanisms.
• Developed applications using #django, #streamlit to facilitate citizen data scientists use our ML solutions.
• Filed disclosure and submitted an internal paper on a filtering technique based on comparative analysis.

Mar 2020 – Jun 2021 · 1 yr 3 mos · Hyderabad, Telangana, India

• Aim to make the car insurance claim process touchless.
• Reduce the time for a complete claim process from a typical 13 days to a mere 5 minutes.
• Enhanced the data extraction and data pre-processing capabilities using multiprocessing techniques.
• I Worked on boilerplate code to streamline the training process for machine learning and deep learning models. This also ensured that the data processing techniques were prone to fewer errors.
• Developed damage detection models using computer vision and deep learning techniques.
• Also responsible for analysing the model capability to generalise and evaluate robustness, the impact analysis of using data processed through different processes.
• Technical details:
• DVC for data version control
• Bitbucket for code version control
• Docker for packaging the code and model
• Neptune for tracking experiments
• Python for development
• Jupyter notebooks for analysis
• Damage classification models
• Efficient Net
• VGG16
• Inception net
• Damage localisation models
• YOLO
• M-RCNN

Oct 2018 – Mar 2020 · 1 yr 5 mos · Hyderabad, Telangana, India

• B2B platform which incentivizes adoption by every organization in an economic network so as to reduce transaction costs & improve market efficiencies.
• Developed a python package capable of modelling agent behaviour based on probability distributions, which can be used to simulate the behaviour of agents in a B2B environment.
• Setup data pipeline for data modeling, complex event processing, blockchain performance analysis and analytics pipeline integration with blockchain.
• Technical details:
• Agent Based Simulation
• Simulate the activity of the participants (agents) in a supply chain environment.
• Interact with the chain code and perform actions on behalf of the agents.
• Configurable agent behaviour (based on sampling distributions.)
• Streaming pipeline
• Stream events from Kafka using spark streaming (pyspark)
• Event parsing and data modeling on streaming events.
• Complex event processing on the parsed data including time based aggregations.
• Custom notifications based on the events.
• Development and testing was done in the cloud computing platform of aws.
• Tools : Python, pyspark,kafka, docker.

Sep 2017 – Mar 2020 · 2 yrs 6 mos · Hyderabad, Telangana, India

• Mentored 2000+ students in ML, AI, NLP, Python and Visualisations.
• Responsibilities:
• Consulting and research projects in the field of Machine Learning, Deep Learning and Optimization.
• Data collection, cleaning and pre processing, visualization and generating actionable insights.
• Implementing algorithms and evaluating the performance gains over present systems.
• Delivered multiple trainings on Machine learning, deep learning and visualization.
• Machine Learning and Deep learning skills: Linear Regression, Logistic Regression, Time Series, KNN, Clustering, Decision Trees, Ensemble Models (Random Forests), Neural Networks, Deep Neural Networks, Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Long Short term Memory (LSTM).
• Tools : R, Python, spark, pyspark, tableau, power bi.
• Packages: plotly, seaborn, scikit-learn, keras, tensorflow.
• http://insights.insofe.com/2019/03/27/music-style-transfer-using-deep-learning-part-1/

Feb 2017 – Sep 2017 · 7 mos · Hyderabad Area, India

• Developed machine learning models to predict insurance claim denial, achieved 86% recall, which helps sutherland to anticipate the possible reasons for claim denial and navigate the claim filing process smoothly.
• Analyze patterns in payments and building prediction models using Deep Learning techniques on R.
• Worked on predicting the propensity to pay and opportunities to prioritize invoices.
• Developed a semantic slot filling mechanism using text mining and Natural Language Processing systems to convert natural language queries into SQL queries.
• Tools : R, Python, spark, pyspark, Tableau, Power Bi

Oct 2016 – Jan 2017 · 3 mos · Hyderabad, Telangana

• Did a data science course from International School of Engineering (INSOFE), Hyderabad to aid my transition into the field.

Sep 2015 – Oct 2016 · 1 yr 1 mo

• Mentoring teams in Transaction Risk investigation

Aug 2014 – Aug 2015 · 1 yr

• Analyzing fraud risk based on attributes of a customer account.
• Pattern recognition and Modus Operandi Identification.