Balasubramanian Swaminathan

Apr 2002 – Jan 2003: Ramco Optima

Migrated a generic process-industry expert-system into VC++/MFC6 with MS SQL and multithreading, enhancing UI functionality and memory robustness; validated real-time control loops on-site at Madras Cement Ltd.

Feb 2003 – Sep 2004: Data Modeling & Analysis Tool

Built a real-time soft-sensor in VC++/MFC using neural networks, PLS regression, clustering and PCA for data cleaning and dimensionality reduction; integrated OPC/SCADA data feeds and deployed for fineness prediction at Ambuja and MCL cement plants.

Feb 2005 – Aug 2005: Optima Enhancement

Led a 4-developer team to extend the Optima expert system with string/date-time data types, port it to 64-bit Itanium, and implement multithreading optimizations for sub-second cement and power-plant loop control.

Apr 2006 – Sep 2006: Genetic Programming Module

Architected a symbolic-regression soft-sensor tool in VC++/C#/VB .NET, leveraging genetic programming to auto-derive mathematical models (e.g., CO-emission prediction) for thermal power-plant analytics.

Jan 2007 – Apr 2007: Rule Engine Evaluation

Evaluated ILOG and JBossRule engines for enterprise lifecycle support, syntax and performance; wrapped rule services in Java/C# web APIs and integrated them into the Ramco Enterprise Suite.

Jun 2007 – Sep 2007: Business Process Optimization Editor & Solver

Directed the creation of a generic LP-solver with an ASCII-to-LP parser and embedded math editor, exposing solver functionality as RESTful web services within Ramco VirtualWorks for dynamic process automation.

Jan 2008 – Jun 2008: Route Optimization for Logistics

Led development of a C++/COM/SQL-based LP solver for the Vehicle Routing Problem with Time Windows, integrating it into Ramco VirtualWorks to deliver enterprise-grade logistics optimization.

Jun 2008 – Mar 2009: RIOS Re-Platform (UHG International)

Led phase-2 re-platforming of RIOS from SQR/UNIX to a .NET/Crystal Reports solution on MS SQL 2005 and DB2; architected the data-access layer in C#, enforced coding standards, and delivered scalable, maintainable reporting services.

Apr 2009 – Dec 2011: MMS Data Management (Molina Medicaid – Maine & Idaho)

Directed a 16-member team to develop and maintain SSIS-based ETL workflows and VB .NET interfaces for Medicaid eligibility, enrollment, and claims data integration into QNXT. Key contributions:

• Designed and automated daily ETL pipelines, eliminating manual processing and enhancing data reliability.
• Developed and optimized SSRS reports by migrating legacy TR/CR reports and enabling real-time operational dashboards.
• Established unit-testing, code-review, and mentoring processes to ensure high code quality and on-time delivery.
• Managed full SDLC—requirements, design, implementation, testing, deployment, and 24×7 production support—tracking change requests and communicating risks and milestones to stakeholders.
• Enforced HIPAA-compliant data governance through rigorous validation, audit trails, and performance tuning.

Jan 2012 – Jun 2012: Facets MSM ICD-10 Remediation (WellPoint)

Led a 10-member team through the ICD-10 upgrade of Cognizant MSM on UNIX; coordinated capacity planning, performance tuning of Sybase/Facets interfaces, and stakeholder communication to meet regulatory deadlines.

Jun 2012 – Dec 2012: CSP F2F Migration (UHG Facets Solution Group)

Managed end-to-end migration of customer-service modules into Facets using Sybase 15, IBM DataStage ETL, and MS Project; implemented risk-mitigation strategies, tracked milestones, and orchestrated cross-functional teams to achieve on-schedule go-live.

Jan 2013 – Dec 2015: Asset Projection & ETL Lead (Kamakura Risk Manager)

• Managed BAU support for Kamakura Risk Manager COTS—ingesting investment, market and interest-rate scenario feeds into SQL Server for CFT, ALM and RBC projections.
• Designed and implemented SSIS-based ETL pipelines sourcing from MarkIT EDM and legacy systems, improving data quality and reducing manual reconciliation by 40%.
• Validated, transformed and loaded security-master, portfolio, loan and derivatives data into the KRM modeling schema, ensuring end-to-end data integrity.

Jan 2015 – Dec 2017: Data Warehouse Performance & Migration Lead

• Migrated KRM's SQL Server asset-projection database to Microsoft Parallel Data Warehouse (PDW), achieving 5× query performance gains for large-scale HPC-grid runs.
• Continued BAU support for Cash Flow Testing and GAAP Loss Recognition—tuning PDW distribution and indexing strategies for sub-minute report generation.
• Led the transition of asset-allocation data feeds from mainframe to PDW, standardizing metadata and implementing automated load processes.

Jan 2017 – Dec 2019: Strategic Migration Lead (FIS Prophet ALS)

• Spearheaded the replacement of KRM with FIS Prophet ALS for asset projections—oversaw Data Conversion System (DCS) pipelines to generate Model Point Files from existing investment feeds.
• Integrated Prophet's encrypted output via the R2DB tool into SQL Server, preserving existing asset-projection formats and enabling seamless business continuity.
• Provisioned and configured the HPC-grid run environment for Prophet ALS, coordinating with infrastructure teams to validate end-to-end workflow.

Jan 2018 – Dec 2020: Cloud Platform Modernization – Azure Asset Data Repository

• Defined functional requirements and oversaw migration of on-prem SQL repositories to an Azure-based Asset Data Repository, leveraging Azure Data Lake and Databricks/Spark for scalable analytics.
• Collaborated with cloud architects to design data ingestion patterns, security policies and CI/CD pipelines—accelerating data availability for downstream risk and finance teams.

1. End‑to‑End Test Strategy & Roadmap

• Mapped > 90% of actuarial "money flows" across Cash‑Flow Testing, LDTI, IFRS & GAAP modules.
• Introduced risk‑based test design that focuses automation where financial exposure is highest.

2. Scalable, Reusable TOSCA Framework

• Built a plug‑in architecture (API, UI, Databricks, file‑transfer) with 80% component reuse across product lines.
• Added data‑driven libraries and self‑healing locators, driving maintenance effort down 45% YoY.

3. CI/CD & DevOps Integration

• Wired TOSCA into Azure DevOps pipelines: every PR spins up a disposable test environment, executes smoke + impact‑based regression, and posts results back to Git as status checks.
• Containerized TOSCA Distributed Execution Agents (DEX) for frictionless scale‑out on Kubernetes/HPC nodes.

4. Databricks Data‑Integrity Automation

• Created PySpark‑backed TOSCA modules that validate Delta‑Lake transformations against actuarial business rules, catching silent data drifts 48 hrs earlier than prior manual spot checks.

5. Actionable Quality Analytics

• Defined a "single pane of glass" Power BI dashboard: coverage, pass‑rate trends, flakiness index, MTTR, and escape defect density.
• Metrics now feed quarterly OKRs and have already supported a 30% drop in escaped production defects.

6. Regression Suite at Scale

• Automated 600+ high‑risk scenarios spanning UI, REST, file‑transfer and grid‑compute workflows; nightly run completes in ≈ 3 hrs (vs 16 hrs manual) with fully automated triage hooks.
• Demonstrated ROI within two sprints: first full suite uncovered a $5M reserve‑calculation defect before UAT.

7. Leadership & Collaboration

• Coaches scrum teams on BDD, API virtualization and shift‑left peer testing; onboarded new hires with a two‑week "TOSCA boot camp."
• Works daily with solution architects, actuaries and DevOps to align pipelines, environments and release cadences.