SUMMARY:

• An experienced Data Scientist with 9 years of experience with over 7 years of expertise in data driven research including data collection, cleaning, preprocessing and deriving key insights for product development, optimization and quality improvement. Proven ability in problem solving skills in the field of emerging technologies like flexible electronics & deep learning physics algorithms and bleeding technologies like quantum computing. An Engineering graduate with two master’s degree in nanoscience and data analytics and multiple s in data science and deep learning.

TECHNICAL SKILLS:

Domain expertise: Advanced semiconductor technology, materials science, nanotechnology, physics, quantum mechanics, machine learning and data science.

Specialization: R&D, productionalization of ideas, process optimization, operations management and product development.

Experience: Data pre - processing, data mining, feature engineering, exploratory data analysis, sentiment analysis.

Data Analytics tools: Tableau, PostgreSQL, Cloud computing in AWS ML: S3, EC2 and Redshift.

Machine Learning: Linear & non-linear models, Natural Language Processing, ANN, CNN, RNN.

Python libraries: TensorFlow, Keras, Scikit, Numpy, Pandas, Matplotlib, NLTK.

R libraries: Syuzhet, ggplot2, H2O, Caret, corrplot.

Writing skills: Eight years of expertise in technical paper writing, documentation and presentation in meetings, conferences and journals.

Business skills: Understanding requirements of non-technical teams, formulating problem statements and translating it to the technical team for implementation.

Leadership: Cross-functional team building in start-up and academic environment with a record of efficient negotiation within groups.

PROFESSIONAL EXPERIENCE:

Confidential, PA

Machine Learning Researcher

Tools: and techniques used: R, Python, TensorFlow, NLP, Text Analytics, Hypothesis Testing, Unsupervised Learning, Supervised Learning, PCA, Sentiment Analysis, Machine Learning, Exploratory Data Analysis

Responsibilities:

• Defined the problem statement and explored the possibilities of using deep learning algorithms for solution identification in Higgs Boson simulated dataset.
• Identified the possibility of using neural networks to isolate and detect new particle signals from background noise through literature analysis and consulting with
• Chief Programmer at LHC, CERN Switzerland.
• Performed statistical analysis on Higgs Boson dataset using correlation plots and found multicollinearity among the feature variables.
• Using table plots, identified four feature variables critical in differentiating signals versus background noise.
• Developed deep neural network using H2O package in R with 5 hidden layers and predicted particle signals with 70 percent accuracy.
• Performed hyperparameter tuning with node optimization and activation functions (ReLU & tanh) and found the most accurate prediction.
• Developed a deep neural net using TensorFlow in Python for detection of exotic particle signals from background noise on a Higgs Boson simulated dataset with test set accuracy of 70 percent.
• Performed data pre-processing and built an artificial neural network model with 6 hidden layers and ReLU activation function using Keras framework in Python for a bank dataset.
• Predicted the ratio of incoming and outgoing customers, with a model accuracy of 85 percent.
• Analyzed the demographic distribution of bank’s dataset in the UK and prepared a dashboard for visualization of segregation of customers in terms of geographies, age, gender and job classification.
• Formulated a problem statement on impact of immigration on the crime rates in the US communities.
• Performed data transformation using square root and box-cox transformation methods.
• Developed multilinear regression model on crime dataset and calculated t-values, p-values, R-squared and adjusted R-squared values among response and predictor variables.
• Found negative relationship between influx of immigrants and increase in crime rates in the US communities.
• Developed an Generative Bag of Words NLP model using NLTK library in Python for a classification problem of predicting positive and negative reviews for a restaurant reviews database with 75 percent accuracy.
• The steps involved text analysis, word2vec modelling, stemming, tokenization and model fitting using Naïve Bayes model.

Confidential, NY

Lead Process Engineer

Tools: and Techniques used: MS Excel, R, Python, NLP, Optical Modelling, Exploratory Data Analysis, Predictive Analysis, Operations Management, Documentation

Responsibilities:

• Developed an NLP model using Syuzhet package in R to compare the emotional variability of William Shakespeare’s novels.
• Compared the emotions with AFINN and NRC libraries and identified the most common emotion and its variance over time.
• Calculated lexical diversity, positive and negative ratings of speeches of present and former first ladies of the US using VADER module from NLTK and performed visual representations of these ratings using Gensim and matplotlib in Python.
• Identified the scope of research on parameters influencing energy consumption of individual houses with the help of Nest sensor.
• Performed hypothesis testing and developed a linear regression model in R to explore the critical variables associated with energy consumption.
• Modified wafer thickness measurement models for new products by identifying outliers and false positives and achieved 100 percent model accuracy for measurements of resistive memory devices.
• Led the process optimization of wafers for 3-D integration and quantum processors, through accurate measurements, identifying outliers and quantitative data analysis and obtained 99 percent reproducibility rate within 6 months of operation and implementation.
• Responsible for documentation and modification of process flow charts and recipes at Materials Research Laboratory.
• Mentored and trained new employees on research projects and equipment troubleshooting and helped them getting assimilated into the company culture more effectively.
• Supervised waste and expired stocks in the labs and coordinated with waste removal team which resulted in cost cutting on future inventories.

Confidential, Albany, NY

Research Consultant

Tools: and Techniques used: MS Excel, Problem Identification, Graph Analytics, Data Generation, Spectral Modelling, Predictive Analysis

Responsibilities:

• Designed process improvement projects for a cutting-edge battery technology development and performed experimentation, data collection, data pre-processing, data visualization and qualitative analysis, and increased the life-time of Li-ion batteries by 40 percent in 6 months.
• Facilitated the process development steps through quantitative data analysis of spectral models obtained during spectroscopic experiments.
• Led the company’s vision in process reproducibility and innovation and initiated the prototype to production process.

Confidential, Albany, NY

Graduate Scientist

Tools: and Techniques used: Problem Identification, Weibull ++, Origin, Curve Fitting, Hypothesis Testing, Data Generation, Inferential Analysis

Responsibilities:

• Identified the product failure problem associated with transistors and conducted frequent meetings with device manufacturer, lab co-workers, other tool owners and device characterization experts, arriving with a way to design experiments for solving failure issues in new generation semiconductor products.
• Identified the reliability problem associated with new generation semiconductor devices and drafted a problem statement followed by experimentation with novel parametric control and techniques.
• Conducted experiments for electrical and reliability characterization and gathered data, performed data cleaning, feature selection and quantitative analysis for hypothesis testing.
• Found the physical mechanism of product failure by formulating new constants to fit experimental data into reliability models and interpreting reliability physics using data visualization.
• Coordinated with technical team and client and liaised them to come with an action plan for joint research project on solder encapsulation.

Confidential, Ruston, LA

Graduate Research Assistant

Tools: and Techniques used: Productionalization, Hypothesis Testing, Problem Identification and Exploration

Responsibilities:

• Assisted professors in supporting 80-students in organic and physical chemistry labs, assisting them in experimentation, identifying issues in the laboratory and finding resolution for each problem along the way.
• Assisted students in assessing potential risks in the experimental process, informing them of potential problems and dangers, while teaching them how to manage these problems in avoiding problematic situations and minimizing risk in the laboratory space.
• Developed a technique from white paper to enhance the functionality of gold nanoparticles, which finds its applications in medical imaging technology.

Confidential

Project Research

Skills used: Project Development, Data Collection, Data Visualization, Data Analysis, Patent Application, Research Paper Draft and Publication

Responsibilities:

• Designed and developed an analytical and chemical technique for extraction, purification and characterization of bioactive compounds from coconut extract.
• Applied for a patent and published two research papers for discovery of bioactive compounds from waste products of coconut.
• Demonstrated leadership and management skills, representing class in department for 3 consecutive years, advocating for student needs and concerns.