PLM (Product Lifecycle Management) Bot is a Chat Bot and developed with Microsoft LUIS deployed in with Azure for a particular use case of vitality score calculation and updated to help user to get the instructions to update through SAP PLM system and also to update the user from the chat bot end through SAP NWBC application. Hence, providing instructions to the user from the Virtual Agent. Integrated SAP SOAP Web-service in the chat bot to update

Environment/Tools & Technologies: C#, Azure, Microsoft Bot Framework, Microsoft Cognitive Services (LUIS), NLP, JIIRA, Visual Studio, SQL Server
Description: Designed, Developed Natural Language Modelling for a virtual agent and integrated with Azure along with other Cognitive Services. Design, Developed ML Model and Language Understanding Model and trained and published in order to get the best accuracy. Deployed in Azure and AWS environments. Spearheaded a team in delivering 2 different use cases to the business team. Productionizing the chat bot in client Azure cloud platform. Deployed the relative models on Azure and implemented with Bot Framework along with a team.

Built, configured, and deployed ML models for automation of customer related services of Test Data Automation Platform, AIOPS in customer production environment. Created seed templates, micro services, client proprietary tools (CMLP), CMLP Studio, docker, Jenkins, h2o, Client SDK in R creation. Designed Model Performance SDK for supporting regression/clustering/classification (single/multi label) performance metrics for customers worldwide. Implemented R Seed templates, created nurseries and worked on video optimizer algorithm and predictive models, worked on Trace Analytics use case and anomaly detection, customer type, service type, prediction and other use cases. Performed Anomaly Detection on Elastic search (ELK) use case & Kibana Logs Data and created predictive service, Trends and Seasonality Time Series Analysis and different use-cases. Created workflow in NIFI for data extraction and data loading from Kafka Topic and retrain models on it.

Designed, Developed Natural Language Modelling for a virtual agent and integrated with Azure along with other Cognitive Services. Design, Developed ML Model and Language Understanding Model and trained and published in order to get the best accuracy. Deployed in Azure and AWS environments. Managed team in delivering 6 different use cases to the business team and has filed a patent for that. Developed, designed and spearheaded a team in delivering different POC of clients like Travelers, Reliance, Triag, DuckCreek, and so on. Deployed the relative models on Azure and implemented OpenAI architecture while spearheading team.

Currently, working on building Knowledge Graph , created entity relation pair from the COVID Analysis data .

Delivered multiple automated Scalable AI solutions for 12 use cases including test-cases automation, anomaly detection, defect classification, Time Series forecasting, and so on.

The objective was to build a custom NER model for B2B platform based on graph DB ontology. Used Solr Search for the free text search, Built more than 4000 training patterns with RASA-NLU and rule based pattern matching to train NER custom model with Spacy. The results from SOlr and spacy were then searched in graphDB.

Built multi -model approach of demand forecasting and Inventory Optimization using Python based stacked Deep Learning algorithms.

Integrated MSLUIS, Bot Framework to build a virtual agent for 6 different use cases for the clients like Travelers, Reliance, Triag, DuckCreek , etc. Unilever PLM Bot was developed using LUIS and it was integrated with SAP NWBC application to automate the task.

Developed Speech to Text Conversion tool with CMU Sphinx and benchmarked with IBM Watson.