Data Architect Analyst

SURYA DHULLIPALLA, PH.D.

SENIOR DATA MODELER/SENIOR DATA ANALYST/SENIOR DATA ARCHITECT

EMAIL: ********@*****.***

cell: 919-***-****

PROFESSIONAL SUMMARY

● 15 years of healthcare experience in analysis, design and implementation of complex healthcare data integration projects and played multiple roles as a Senior Data Analyst, Senior Data Modeler, Senior Data Architect, Senior Technical Engineer.

● Involved in the analysis and design of the data models - a development blueprint that will serve as the foundation for the ongoing development of data platform representing curated and integrated Pharmacy Data

● Key designer of the IBM Unified Data Model for Healthcare [UDMH] and Flexible Analytics - an industry model for Providers [hospitals] and Payors [insurance companies]. The data warehouse model includes Provider [data elements based on HL7/FHIR Resource] and Payor [data elements based on HIPAA 837-I,837-p, 835] where the Longitudinal Patient Record (patient’s one source of truth record) resides in this model. The transactional data model is designed based on the Data Vault design pattern and the dimensional data model based on Kimball’s design approach. Please see :https://www.ibm.com/docs/en/udmh

● Worked with Subject Matter Experts to obtain requirements for the design of analytic models. Provided training datasets to Data Scientists to build and test the analytic models that predict resource and utilization, patients costs, complications, readmissions and mortality.

● IBM Micromedex Watson Cognitive Platform is the web portal used by pharmacists worldwide. Designed the underlying ontology model that the Micromedex web portal sits on. To build the Drug Ontology Model, collaborated with Pharmacists to do intense data analysis on medications and define concepts, data properties and object properties for Drug, Drug Interactions, Drug Food Interactions, Drug Ethanol Interactions, Drug Allergy, Pharmacokinetics, Indication, Contraindication, Black Box Warning, Precautions, Adverse Effects, Dosage, Dose Adjustment, Dose Administration, Clinical Condition, Finding, Ingredients, Drug Brand, Drug Class, Mechanism of Action, Monitoring, Efficacy, Pregnancy Lactation, Regulatory Status. Trained Watson with healthcare and life sciences data that uses natural language processing to understand contextual meanings of the clinical data. Please see:

https://www.ibm.com/in-en/products/micromedex-with-watson

● The Benefits Annotator web application allows the Benefits consultants to get the information about cost share data elements- Coinsurance, Copay, Deductible amounts that the member pays or the provider pays for a given service. The application allows the user to upload the 200 page PDF Summary Plan Documents and Summary Benefits Coverage documents that are converted into the JSON format by the OCR engine and in turn calls the ACD Benefits cartridge that uses Natural Language Processing to return the coinsurance, copay, deductible values for the services obtained by the member. Was a Key Designer of the Benefits Annotator Cartridge by collaborating with Benefits consultants to model the cost share data elements for services such as: Primary Care Visit, Specialist Visit, Preventive Care, Screening, Immunization, Imaging Maternity, Mail Order/Retail Order for Generic and Branded Drugs, Specialty Drugs, Inpatient Facility and Professional Fee, Outpatient Ambulatory and Professional Fee, Emergency Services, Urgent Care, Behavioral Health-Inpatient and Outpatient Services (Mental Health, Substance Abuse), Dental and Vision – Eye Exam, Prescription Glasses for the in-network and out-of- network.

● Led various enterprise-wide data warehouse implementations at hospitals (UNC Chapel Hill, Baylor, Duke) across multiple source systems that belong to multiple tenants. Worked with Medical Directors, Physicians, Nurse Practitioners, Clinical Coordinators, and Statisticians on designing and implementing the data warehouse and several web applications based on J2EE Engaged with stake holders to build roadmaps and review Functional Requirements Documents to determine scope and deliverables of application. Translated both Business and Functional Requirements Documents into Technical Requirements Documents.

● Interacted with Subject Matter Experts and Data Analysts at various client locations. Was in an advisory role in helping the data analysts in their source data analysis and mappings and creating business rules for cleansing data.

● Worked with Business Intelligence team to design dashboards/ reports for data access against datamarts.

● Worked on projects following Agile process using Rally.

● Worked with data governance groups who represented a cross-section of stakeholder groups, data stewards to establish standards, requirements, guidelines, data definitions for the data.

● Experience on the design of Object Models, Kimball, Inmon, Data vault design methodologies and design of clinical and operational ontologies.

Technologies

Databases/Queries: DB2, Oracle, SQL Server, Microsoft Access, SQL, PL/SQL, Hadoop/HIVE-Big Data Data Modeling Tools – Infosphere Data Architect, Erwin, ER Studio, Toad, Visio Reporting Tool: COGNOS

Web ontology Tool (OWL): Protégé 5.2, RDF/XML

Languages/J2EE Architecture: JAVA, JDBC, SQL, PL/SQL, Python Cloud: Azure, Snowflake

Data Format: XML, JSON

Integration: HL7, FHIR, EPIC [Healthcare], Allscripts, DBMotion, VISION. Code Management Tools: GIT, SVN

Methodologies: ER Models, Kimball, Inmon and Data Vault data warehouse design, Object Modeling

– Rational Unified Process

PROFESSIONAL EXPERIENCE

July 2021 – Feb 2022 United Health Group-Optum Labs, Durham, North Carolina Senior Ontology Engineer

● Worked as a Senior Ontology Engineer - involved in the analysis and design of ontology models underlying the Individual Health Record

Jan 2021 –June 2021 CareSource-Dayton, Ohio)

Principal Healthcare Data Architect/Data Modeler

● Worked as a Senior Data Architect/Data Modeler - involved in design and development of data models that support use cases which require processing of data from Express Scripts into the Data Lake, and landing in a data platform representing curated and integrated Pharmacy Data Jun 2020 – Jan 2021 United Health Group-Optum, Durham, North Carolina Principal Healthcare Data Architect/Data Modeler

● Worked as a Senior Data Architect - wrote groovy shell scripts to migrate claims data files from Datalake to Microsoft AzureBlob containers and finally into Snowflake tables on Azure cloud.

● Collaborated with the Data Analysis team within US and Off-shore to analyse clinical and claims data from different clinical data organizations to design and create Azure containers, Snowflake objects- Stage tables, Storage Integration, Tables, Pipes, Streams and Tasks. Wrote SnowSQL scripts to load data from Azure blobs to Snowflake tables.

● Kafka is an open source software which provides a framework for storing, reading and analysing streaming data. Involved in creating mappers using Go language and Golang libraries to read the raw kafka topics to output the canonical models based on avro schema that adheres to the enterprise logical model standards as per the Data Governance.

● The canonical models that are the output kafka topics are consumed by the downstream applications which eventually are stored on the Snowflake data warehouse on Azure cloud.

● The canonical models are validated for its accuracy such that the actual input data and expected canonical data match completely. The data formats used are avro/json Oct 2016 – Jun 2020 IBM Watson Health, Durham, North Carolina Principal Healthcare Data Architect/Data Modeler

● Extraction of the provider and payers’ claim data elements using SQL queries

● Migration of Financial and Clinical Data to Watson Health VBC platform.

● Design of Flexible Healthcare Data Analytics Data Model version 1.9 [Truven Methodologies] plugin to the IBM Universal Data Model for Healthcare version 9.2.

● Building ontologies for the Watson conversation assistant/chat bot service using Protégé 5.2. Reverse engineering the complex ER models to the ontology models.

● Key Designer of the Benefits Annotator Cartridge that uses Natural Language Processing to return the coinsurance, copay and deductible amounts from the provider and member centric perspectives. Migration of Financial and Clinical Data to Watson Health VBC (Jan 2018-Jan 2020)

● Enterprise wide migration of Patients Encounter data and Payer claims data to IBM Universal Data Model for Healthcare v9.3 The primary users of this data are CEO, CMO, and CFO for viewing the data on the respective dashboards. Collaborating with executives, project managers to create road maps for the following features.

● Physician and Program Performance: Track patient-level claim and clinical measures to assess individual physician and overall VBC program performance against targets.

● Facility and System Performance: Track hospital, ambulatory, and post-acute quality outcomes at the encounter level, that includes operational and financial performance

● The Financial KPIs such as the Operating Margin %, Total Expense, Total Labor Expense, Supply Expense, Drug Expense and Paid Hours per new discharge volume will provide Financial insights to the CFO at System and Facility level.

● The Clinical KPIs such as Risk Adjusted Mortality Index, Expected Complications Rate Index, Risk Adjusted Expected Readmissions, Length of Stay, Expected Cost of Care from Expected Resource Demand Model, Hospital Acquired Conditions provide clinical insights to the CMO at System and Facility level.

● The technologies include Oracle, Tableau, Infosphere Data Architect for the implementation. Design of the UDMH v9.2 Flexible Analytics v1.8 Model (Oct 2016 – Jan 2018)

● Designed data model that supports analytics for both Provider and Payer including: Admissions, Cost of Care-Commercial Model, Cost of Care- Medicaid Model, Disease Staging, Medical Episode Grouper, Outpatient Event Grouper, Service Categories and Quality Measures Engine, Expected Complications Rating Index [ECRI], Risk Adjusted Mortality Index [RAMI], Ratio of Cost to Charge [RCC], Expected Resource Demand/Patient Financial Data [ERD/PFD] using Kimball's Design methodology.

● The Admissions methodology gives analysts a tool for assessing the details of inpatient-admission utilization and costs from Medical Claims and Pharmacy Claims.

● Cost of Care Methodology helps improve health and manage costs by understanding resource utilization from the Administrative Claims. The commercial cost of care model predicts the risk of healthcare resource consumption for commercial population.

● Disease Staging methodology uses Medical Claims to categorize patient diagnoses into one of 580+ disease categories and one of four severity stages based on age, sex, principal diagnosis, procedures performed and is an input to group multiple events to an episode and assign risk scores to a Risk Adjusted Cost of Care Models.

● Medical Episode Grouper with Cost of Care Model uses claim records to generate a risk-adjustment model that leverages the patient illness burden to adjust the expected cost of an episode.

● Outpatient Event methodology groups multiple claim records from multiple providers for a single event. Service Category methodology assigns all services from facility, professional, drug claims, and non-claim payments such as capitation amounts and premium payments, to a standard set of categories.

● Expected Complications Rate Index (ECRI) provides a method to identify risk-adjusted complications rate to allow fair comparisons among healthcare organizations.

● Risk Adjusted Mortality Index [RAMI] methodology computes the risk of death for hospital patients based on clinical and hospital characteristics- geographic location, size and teaching status, specialty type and urban or rural setting.

● Ratio of Cost to Charge [RCC] methodology estimates patient costs based on the patient charge information.

● Worked with Subject Matter Experts to obtain requirements for the design of analytic models.

● Worked with ETL developers to build queries for data preparation and provided training datasets to Data Scientists to build and test the analytic models written in Python Nov 2011 – Sep 2016 Perficient, Inc.

Senior Technical Architect

CLIENT: Baylor Scott & White Health System, Dallas, Texas

● Supervisory Role in the Enterprise wide data migrations across the health system.

● Collaborated with SMEs to perform detailed data analysis of the source systems - Epic, Centricity, Vision, DBMotion. The data elements included Provider and Payer Claims data (837I, 837P and 835) as well as the clinical data.

● Created Source/Target Mapping specifications between different home-grown legacy systems in Oracle, SQL Server databases and the target DB2 database that is based on the IBM Universal Data warehouse Model for Healthcare (UDMH).

● Created table SQL scripts to the DBA to implement the database.

● Provided SQL query scripts to the ETL team for migrating data from the source systems to the DB2 database.

● Worked with the Business Intelligence team in the design of dashboard that displays the clinical KPIs in COGNOS by providing the necessary queries to identify the table data.

● Worked with the testing team in the creation of test cases against the identified use cases.

● The technologies involved in the migration are: Infosphere Data Architect for Data Modeling, Data Stage for ETL loads and COGNOS for Dashboard Reports.

CLIENT: University of North Carolina Hospitals

● Supervisory Role in the Enterprise wide data migrations across the health system.

● Collaborated with SMEs to perform detailed data analysis of the patient’s clinical data in the legacy source systems as well as the EPIC system.

● Created Source/Target Mapping specifications between the home-grown legacy tables in Oracle database and the target columns of the DB2 database that is based on the IBM Universal Data warehouse Model for Healthcare (UDMH).

● Created Source/Target Mapping specifications between the EPIC tables/columns and the target IBM Universal Data warehouse Model for Healthcare (UDMH) columns.

● Provided create table SQL scripts to the DBA to implement the database.

● Provided SQL query scripts to the ETL team for migrating data from the source systems to the UDMH data model implemented in DB2 database.

● Worked with the Business Intelligence team in the design of dashboard that displays the clinical KPIs in COGNOS by providing the necessary queries to identify the table data.

● Worked with the testing team in the creation of test cases against the identified use cases.

● The technologies involved in the migration are: Infosphere Data Architect for Data Modeling, Data Stage for ETL loads and COGNOS for Dashboard Reports.

CLIENT: Premier Healthcare Inc

● Key Architect/Senior Data modeler in the design of the INDUSTRY model – Unified Data Model for Healthcare [UDMH] sold to several hospitals in United States

● Designed the IBM Healthcare Product Data Model [HCPDM] Version 8.6 as a key member of the Joint Data Modeling Team [IBM, Premier and Perficient] using IBM Infosphere Data Architect Version 9.1

● Worked with Subject Matter Experts from Premier as well as IBM to review the Use Cases and HL7 specifications to design the data models for different subject areas.

● Designed the transactional model based on Data Vault methodology using and Dimensional Warehouse Model based on Kimball’s methodology- Star Schema approach.

● Used IBM Infosphere Data Architect, data modeling tool to design conceptual, logical and physical models in subject areas: Encounter, Care Management, Patient Medication, Clinical Order, Procedures, Diagnosis, Claims, Genomics, Surgery, Schedule and Appointment, Labor, Party and Role [Organization, Practitioner, Person, Patient, Tenant and Contact], Document Management, Unified Standard Codes.

● Designed Business Solution templates using Dimensional modeling approach based on measures obtained from different agencies to report patient care and quality such as: National Committee for Quality Assurance[NCQA]; Agency for Healthcare Research and Quality [AHRQ]; Hospital Quality Alliance (HQA); Centers for Medicare & Medicaid Services (CMS) partnered with the Agency for Healthcare Research and Quality (AHRQ) to develop HCAHPS—the Hospital Consumer Assessment of Healthcare Providers and Systems Survey; Physician Quality Reporting Initiative (PQRI); Inpatient Psychiatric Facility Quality Reporting (IPFQR); Hospital Outpatient Quality Reporting (OQR); Ambulatory Surgical Centers Quality Reporting (ASCQR).

● Designed dimensional models in Clinical Analytics: Chronic Disease Management Analysis; Clinical Care Variability Analysis; Patient Safety Analysis; Pharmacy Utilization Analysis; Surgery Analysis

● Designed dimensional models in Disease Specific Measures: Cancer Screening and Care Analysis; Quality of Care - Heart and Vascular Disease Analysis; Stroke Disease Assessment And Care Analysis

● Designed dimensional models in HEDIS (R) 2013: HEDIS (R) Access/Availability of Care Support; HEDIS

(R) Effectiveness of Care Support; HEDIS (R) Experience of Care Support; HEDIS (R) Health Plan Descriptive Information Support; HEDIS (R) Utilization and Relative Resource Use

● Designed dimensional models in Operational Performance: Inpatient Operational Performance Analysis; Inventory Cost Analysis; Labor Productivity Analysis; Labor Utilization Analysis; Location Utilization Analysis; Medical Equipment Utilization Analysis; Readmission Prediction Analysis; Referral Analysis; Survey Analysis

● Designed dimensional models in Regulatory Reporting: ACO; AHRQ Quality Indicators; Acute Claim Measures; Core Measures; Hospital Acquired Condition Measures; Meaningful Use Analysis; NHSN Statistics

● Designed dimensional models in Service Line Analytics: Operating Room Performance Analysis; Pharmacy Compliance Analysis; Inpatient Service Performance Analysis; Radiology Department Performance Analysis

● Facilitated creation/migration of data objects (DDL) across Development /QA/Production.

● Established Standards, Guidelines, and Best Practices around data modeling effort.

● Conducted Model Reviews with project team members.

● Generated Physical data models on DB2 [normalized models] and IBM Netezza [dimensional models] data warehouse appliance environments.

● Published and communicated models to Business Intelligence and ETL teams. Participated in Data Governance discussions to establish policies and procedures. Effectively communicated with team members, Senior and Executive Management to make strategic recommendations. Dec 2004 – Oct 2011 Duke University Medical Center Senior Systems Engineer

● Designed various data models using the data modeling tool - ERWin 9.2 – to build data warehouse and databases as well as web applications based on J2EE [3-tier architecture- spring framework].

● Led the design of data warehouse, neurological research database, case report form, perioperative genomics research database, neuro cognitive laboratory system, designed and implemented the Charge Voucher – a billing system available in the operating rooms at Duke Hospitals. Analyzed large clinical data sets obtained from perioperative- pre-op, intra-op, post-op and PACU systems – Surgical Information Systems [Oracle], Innovian Anesthesia Systems [SQL Server] for various research projects.

● Neurological Research Database: Led the data migration project that involved the migration of data from legacy systems. Enhanced the data model of a legacy application consisting of over 200 tables for transforming it into web application by establishing additional relationships for improving performance and features. Wrote design specification to map attributes from different source systems, loaded data to Oracle database using SQL Loader utility. Wrote complex SQL queries, inline views, triggers, PL/SQL packages, procedures and functions in Oracle. Experience in Master Data Management- analyzed the metadata of the source systems and mapped source to target tables. Built web application based on the Spring Framework.

● Case Report Form: Created conceptual, logical [ERD modeling] and physical data model from scratch for a paper-based Case Report Form – CRF, [~200 pages, 3000 fields] and collaborated with operations manager and clinical coordinators for developing PeRiGren [Perioperative Genomics Research], a web application. Mapped the logical [ERD modeling] data model to a physical data model and created scripts for a physical design. The ER models were normalized to avoid data redundancies. Built web application based on the Spring Framework.

● Duke Neurocognitive Laboratory System: The application allows neuro testers to score battery of tests that patients participated in. Was involved in the clinical data analysis related to the patient’s neuro degenerative disorders. Interacted with Subject Matter Experts to design the conceptual, logical and physical data model. Created database scripts to implement the database in Oracle 8.0.

● PiReports: The application allows users to enroll patients for clinical trials, investigators to view the enrollments and funding information, medical director, operations managers and business staff to generate reports regarding investigator’s financial and patient’s enrollment data. Was involved in the clinical data analysis, created logical and physical data modes, wrote SQL scripts to create the Oracle database and SQL queries to access the data.

● Freezerworks: Analyzed the data model of the Freezerworks, a sample tracking and freezer inventory program that stores data about the DNA samples which is a 3rd party tool. Mapped the fields of the source and destination tables and assisted the DBA in creation of scripts for analyzing the source data, and ETL scripts in Oracle PL/SQL for the clinical data migration of a million studies. Experience in Master Data Management- analyzed the metadata of the source systems and mapped source to target tables.

● Genomic and Natural Diversity Data: Participated in discussions with Working Groups, Sabbatical and Postdoctoral fellows at National Evolutionary Synthesis Center to create Ontology Models for evolutionary biologists to implement the databases in PostgreSQL and Oracle environments. Designed and implemented the Clinical Application System in Visual Basic 5.0/Oracle at Center for Human Genetics. Wrote ETL scripts in Oracle PL/SQL to load genomic data to Oracle database. Experience in Jun 2004 – Nov 2004 IBM, Research Triangle Park, NC Senior Consultant

● Notebook Finder: Created conceptual, logical [ERD modeling] and physical data model in Data Architect from the requirements specification by collaborating with senior managers for a web application - Notebook Finder. Mapped the logical data model to a physical data model and created scripts for a physical design in DB/2. The application allows users to retrieve configurations, prices of the products and related information available on the ibm.com website. Mapped entity relationship diagrams [ERD] to class diagrams and Data Access Objects – DAO in the JDBC connectivity layer during the development of the web application

Feb 2003 – May 2004 Glaxo SmithKline, Research Triangle Park, NC Senior Consultant

● Clinical Planning and Management Data Migration: Migrated clinical trials data from legacy systems

(Visual Basic/Oracle 8.0, Power builder/Oracle 7.3) to Siebel e-clinical application (Oracle 9.2) and 1 million studies were migrated. Created conceptual, logical [ERD modeling] and physical data model in ERwin from the migration specification by collaborating with senior managers for a Clinical Planning and Management System (CP&M). Analyzed the data models of the legacy clinical and finance systems, spreadsheets and destination Siebel e-Clinical tables. Worked on Master Data Management- analyzed the metadata of the source systems and mapped source to target tables. Created scripts for ETL, error handling, logging and audit functions, PL/SQL packages, stored procedures, functions, triggers, PL/SQL collections, views, synonyms, sequences in Oracle 8i, indexes to optimize the performance of the queries, export and import utilities.

Jun 2002 – Jan 2003 Eli Lilly, Research Triangle Park, NC Senior Consultant

● Global Chemical Sample Archiving/Registration System: Created conceptual, logical and physical data model in Erwin from the requirements specification by collaborating with senior managers for Global Chemical Sample Archive/Registration Technologies Submission and Request Automation system. The web application was implemented using MVC design, spring framework, JSP/JSTL and struts libraries, JDBC connectivity layer, JAVA servlets, Eclipse, Tomcat, MDL-ISIS/Host API, ISIS/Draw, ISIS/Base and Oracle database. Used MDL Chime plugin for structure visualizations in the web applications. Communicated with scientists globally to obtain user requirements, created design specification documents, created data model, mapped data model to object model, created databases, used various MDL- ISIS functions embedded in Oracle queries for retrieving chemical structures and substructures.

● Electronic Lab Notebook System: Automated the paper-based lab notebook. Created conceptual, logical [ERD modeling] and physical data model in Erwin from the requirements specification by collaborating with senior managers for Electronic Lab Notebook system. Identified entities and relationship from the user requirements. Mapped the logical data model to a physical data model and created the database in Oracle.

Dec 1999 – Mar 2002 Caribou Lake, Minneapolis, MN Senior Consultant

● Upgrade of the Hospital Management System: As a Senior Data Analyst and Senior Data Modeler, Database Architect, interacted with Subject Matter Experts(clinicians, nursing staff, physicians and hospital manager) to create a Due Diligence Report that emphasizes the features and the new domains to be included to the existing data models, towards the Upgrade of the Hospital Management System which was implemented in the Oracle database. The system includes clinical, financial and administrative modules such as ADT (Patient -Admission, Discharge, Transfer), Medical Records, Clinical Pathways, Order Entry, Pharmacy, Patient Billing and HIS Common (admin functions). Dec 1997 – Nov 1999 Javelin Technology, Minneapolis, MN Senior Consultant

● Designed various departmental databases such as Employee Database system and Timesheet Database in DB/2.

EDUCATION

Indian Institute of Chemical Technology

Ph.D. Organic Chemistry

University of St. Thomas, Minneapolis MN

M.S Software Engineering

Contact this candidate