Amitabh Priyadarshi

• 706-***-**** • adtwdg@r.postjobfree.com • Atlanta, Georgia, 30328 • LinkedIn Experience

Enterprise Information Technology Services (UGA) Athens, Georgia Graduate Assistant 01-16 to 05-22

• Contributing software engineering expertise in the development of product/feature through Agile Software Lifecycle for UGA OIR.

• Design and developed dashboards and visualization, Change request, Fixing the bugs for proprietary Analytical tools for UGA Institutional Research using Java, Spring Boot, HTML, D3.js, math.js, jStat.js, Ajax, JQuery, RESTful API and JSON.

• Spearheaded sentiment analysis project on twitter data for UGA events, a Visualization dashboard using D3.js, JQuery, NLP and Java. Complex Carbohydrate Research Center (UGA) Athens, Georgia Graduate Research Assistant 09-14 to 05-15

• Collaborate within team and across teams especially with research org to ensure successful integration and landings.

• Active participation in design discussions with researchers and scientists for improving functionalities.

• Investigate issues and bugs to conclude problem root cause and formulate corrective action recommendations.

• Contributed in Qrator a two-stage review process system for curating glycan structures.

• Change request, Fixing the bugs.

Aequor Information Technologies Pune, India

Senior Software Engineer 01-10 to 03-13

• Contributed software engineering expertise in development of product/features through Agile Software Lifecycle, from requirements definition to successful deployment.

• Enriched codebase with Junit and Integration test cases.

• Worked along with product owner, analytics team to develop proprietary ERP System named Alpharetta VIP 7, which is an order processing and product development system for clients involved in design and maintenance for mechanical devices used in mining.

• Involved in sprint planning, backlog grooming, daily scrums, sprint reviews and sprint retrospectives, contributing towards improving team efficiency.

• Provided code feedback to other developers through pull requests. Education

Doctor of Philosophy (PhD) Athens, Georgia

University of Georgia 08-14 to 12-22

• Major in Computer Science, studied Core and Advance level courses with minor in Statistics.

• Research interest in Querying, Data Mining and Machine Learning on Partitioned Knowledge Graphs.

• Ontology and Knowledge Graph Partitioning, Evaluation and Testing. Master of Computer Applications (MCA) Pune, India

Bharati Vidyapeeth University 07-06 to 06-09

• Studied Core and Advance level courses focusing on development of computer applications. Bachelor of Computer Applications (BCA) Delhi, India Indira Gandhi National University 07-01 to 12-04

Technical Skills

Programming Languages: Java, Python

Enterprise Skills: REST, AWS, Spring Boot, MVC, Kubernetes, Dockers, Servlet, Struts2, JPA, Hibernate, JDBC, JBoss, Tomcat Web Development: JavaScript, jQuery, NodeJS, VUE, React, D3, HTML, CSS, bootstrap. Database: MySQL, MS SQL Server, Neo4J, PostgreSQL, DynamoDB, MongoDB, Virtuoso (RDF and OWL). ML Packages: Pandas, NumPy, Matplotlib, Scikit-Learn, SciPy, NetworkX, OpenCV. Machine Learning: Classical ML, Graph Mining, Graph Embeddings, Representation Learning, Reinforcement Learning, NLP. Publications

• A. Priyadarshi and K. J. Kochut, "WawPart: Workload-Aware Partitioning of Knowledge Graphs," Cham, 2021: Springer International Publishing, in Advances and Trends in Artificial Intelligence. Artificial Intelligence Practices, pp. 383-395.

• A. Priyadarshi and K. J. Kochut, "AWAPart: Adaptive Workload-Aware Partitioning Knowledge Graphs," in SEMAPRO 2021, The Fifteenth International Conference on Advances in Semantic Processing, Barcelona, Spain, 2021: Thinkmind Digital Library, 2021, pp. 12-17.

• A. Priyadarshi and K. J. Kochut, "PartKG2Vec: Embedding of Partitioned Knowledge Graphs," in International Conference on Knowledge Science, Engineering and Management, 2022: Springer, pp. 359-370.

• R. Taujale, A. Priyadarshi et al., "GTXplorer: A portal to navigate and visualize the evolutionary information encoded in fold A glycosyltransferases," Glycobiology, vol. 31, no. 11, pp. 1472-1477, 2021, doi: 10.1093/glycob/cwab082. Projects

PartKG2Vec

• Created PartKG2Vec as a Graph embedding method for datamining on partitioned Knowledge graph which transform the nodes of a graph into a set of vectors using Python and its packages (numpy, networkX, genism (word2vec), METIS, PyLucene).

• Created indices, using Apache Lucene, on all triples in the knowledge graph, based on their subjects, predicates, and objects.

• Using this indexing to convert the knowledge graph into a numeric identifiers graph to partition using METIS.

• Embedding generated from the algorithm I created, captures the graph topology, node-to-node relationship, and other relevant information about the graph, its subgraphs, and nodes similar to node2vec and DeepWalk but much faster. AWAPart

• Adaptive partitioning method for large-scale knowledge graphs without replication is the progression of WawPart where a query workload evolves over time (queries are added/deleted), created in JAVA with Jena, Lucene, and Virtuoso as triple store.

• Created module for conversion of different format of knowledge graph into RDF to best fit the application and query against it using Apache Jena.

• Created an algorithm which continually re-partitioning the Knowledge graph by swapping the triples to accommodate changes in the workload. GTXplorer (Funded by NIH)

• Created Glycosyltransferases (GTs) explorer, used to understand the nature and role of carbohydrates and related glycol-enzymes.

• The tree view mode provides an overview of the sequence-structure-function relationships of GT-A families.

• Allows users to select phylogenetically relevant families for comparisons in the alignment view at the residue level using annotated web-logo stacks of the GT-A core specific to the selected clade, family, or subfamily.

• All data are easily accessible and can be downloaded for further analysis in various form like FastaGTADomain, FastaFullSequence. WawPart

• Created a knowledge graph partitioning method based on query workload using JAVA.

• Indexing of Knowledge graph is done for easy partitioning based on subject, object and predicate using Apache Lucene.

• The partitioning is tailored to a given query workload which reduces the distributed joins across queries Using Hierarchical Clustering.

• It improves the overall workload performance when federated queries are queried against the Knowledge Graph stored in Virtuoso. USNews Ranking Simulator for UGA

• A visual analytical tool which simulates UGA ranking based on quantitative and qualitative measures that education experts have proposed as a reliable indicator of academic quality, it is based on research views of what matters in education.

• It helps OIR to compare immediately the relative quality of institutions.

• The indicators include input measures that reflect the school’s student bodies, faculties and resources. It also includes outcome measures that signal how well institutions are engaging and educating their students. Academic Advising Tool

• Created a Dashboard for graduate coordinator to see the student academic information and progress in D3.js and JQuery.

• Analytical tool to advise students to follow one of the paths for completion of their degree faster based on the graduation likelihood score for each path. These different possible paths are visualized as Sankey diagrams.

• Application will reduce valuable time of graduate advisor while advising students because of handiness of all valuable information about student.

• A module ‘what if’ uses predictive analytics to predict a preferred course for that specific student in which he/she can succeed. OIR Connection

• Created an interactive graphic webpage using D3.js to create network map, chord dependency map between departments for teaming.

• Dashboard to show the involvement of team in different departments and to find out which team needed more resources, funds etc.

• I made it as a stand-alone application for portability using electron.js.

• Also created REST based server to service the request from application for generating information’s using JAVA. UGA Sentiments Analysis during Events

• Created a word cloud for UGA sentiments during events using D3.js. which measures how often a sentiment about UGA is mentioned in a tweet.

• The application uses Stanford CoreNLP library java API to analyze tweets extracted by Twitter4J library.

• A Stand-alone application created using electron.js which request for data using webservice provided by REST

• Sentiment analysis is done using JAVA and REST used as a webservice to aid the application. SPIA Student Major Paths

• Creates a webpage to visualize students’ major paths to graduation for School of Public and International Affairs using D3.js.

• Sankey diagram shows students path to graduation at college or major level, first year or transfer student etc. for enrollment year 2009 to 2018.

• Created an analytical tool in JAVA with REST as web service which facilitate with information for insight to the SPIA department into taking decision regarding the retention of the students.

Onto Browser

• Created a web application for browsing Ontologies which is a graphical way to browse the ontologies.

• The algorithm automatically generates a user interface with hierarchical trees and links based on the ontology to browse through the ontology.

• Provided sets of pre-built queries for easy access to the classes, properties, triples etc. of the ontology.

• A set of auto-generated queries are provided to further understand the data and schema of the ontology.

• For advanced level users, there is a query engine which helps users to write their own SPARQL queries against the ontology.

Contact this candidate