LAKSHAY WADHWA

Tech Lead of the AI team and leads the product development team. Handles the technical aspects of the WRU.ai product and leads client interaction for requirements and building the new features. Built a user specific real time recommendation service for the media houses for showing appropriate recommendations to their end users based on their user history to give a personalized experience for the end users. Built a content similarity engine based on word vectors and high level ANN algorithm to recommend similar products and similar news to the current product or current news article user is viewing. Built various NLP services like tagger, summarizer to help clients with SEO. Helped the clients to increase user engagement on their respective platforms using recommendation engines upto 35%. Automated end to end data pipeline using data preparation techniques to prepare data for data analysis. Introduced state of the art technologies in the team and wrote end to end python services for deployment in the production environment.

Integral part of the Conversational AI team and works in the Data and Analytics group in Voice of Customer domain. Interacted with different data vendors to collect data and automate the data flow from vendors to data lake using API calls. Built data relevancy engine to automate the data preprocessing especially textual data that can be used later in analytics. Converting textual data to numerical form using various pretrained embeddings like Glove and word embeddings to prepare data for analysis. Build classification models to recognize noise in the data and remove the unactionable data from sources based on aggregation vs actionable data. Modelling various aspects of textual data like topic modelling, gap enhancement, churn etc. using classification techniques based on Machine Learning techniques like Logistic regression, Naive Bayes, XGBoost, classification trees etc. and Deep Learning techniques like BERT (Bidirectional Encoder Representations from Transformers). Finding patterns in the data using EDA(Exploratory data analysis) and cluster analysis to group similar data to extract insights out of them. Built a VOC(voice of customer) tool to analyze customer feedback and automated the processing pipeline.

Indulged in excessive Exploratory data analysis on drilling and well data using statistical, multiprocessing and visualization packages in python and Spotfire. In depth statistical analysis of present algorithms to improve the products efficiency for more productivity with respect to the given conditions and inputs. Finding patterns in the data to choose which set of equations are more favorable for the final product and building dashboards in Tableau and Spotfire for decision making purposes, so to choose which wells are matching the key performance indicators and if not, what factors are not letting them meet the performance bar. Performing Machine Learning on data to identify the clusters directly and establishing a workflow for identifying the patterns in the data. Managing data using AWS RDS, AWS S3 and MySQL Workbench for deployment of database for the drilling web apps.

Worked on the GE Predix, an industrial cloud based platform (PAAS) to collect industrial data from machines connected to the server. Designed algorithms and flowcharts to create new algorithms for the data analytics and interpreting data, analyzing results using statistical techniques. Developed a product for a Swiss based healthcare giant, a multi million dollar project for product development involving a real time data transfer between servers. Compiling, assessing, wrangling and cleaning data to draw insights out of it. Worked on various Proof of Concepts to research feasibility of various ideas discussed with the leadership team.