AI-Powered MLR Automation Platform
Challenge
Pharmaceutical companies face lengthy, manual review cycles for Marketing, Legal, and Regulatory (MLR) compliance of marketing materials. Reviews can take weeks, requiring multiple expert stakeholders to verify claims against safety information and clinical evidence.
Solution
Designed and delivered a scalable, AI-driven automation platform on AWS that automates MLR review processes. The system:
- Processes text, images, and layout from marketing materials
- Integrates AWS Bedrock LLM models to evaluate claims and graphics
- Validates truthfulness, completeness, fair balance, and compliance
- Generates intermediate reports and final AI-driven recommendations
- Ensures audit readiness and regulatory compliance
Technical Stack:
Cloud: AWS infrastructure (EC2, S3, Lambda) | Core Services: Python microservices | AI/ML: AWS Bedrock LLM models | Architecture: Scalable, resilient, event-driven
Impact
- Dramatically reduced review timelines
- Maintained 100% compliance with pharmaceutical promotion standards
- Improved accuracy and consistency in compliance checks
- Built audit-ready system for regulated environments
Key Learnings
Demonstrated ability to lead AI initiatives in highly regulated industries, integrate LLMs into complex workflows, and drive innovation in compliance automation.
AI-Powered Legal Tech Platform for Medical Documentation Assessment
Challenge
Legal professionals needed an intelligent system to assess sensitive medical documentation for case evaluations. Manual review was time-consuming and subjective, requiring experts to cross-reference multiple documents.
Solution
Built an AI-powered platform that:
- Processes sensitive medical documentation securely
- Integrates ChatGPT LLM for intelligent document analysis
- Provides legal professionals with expert-level insights via conversational interface
- Ensures high availability, security, and compliance for regulated environments
- Streamlines compensation assessments and case evaluations
Technical Stack:
Cloud: Azure infrastructure (Web Apps, SQL Database, Storage) | Backend: .NET for orchestration | AI: ChatGPT LLM integration | Data Pipeline: Documentation preprocessing | Security: HIPAA/compliance-ready
Impact
- Reduced case processing time significantly
- Improved case accuracy and consistency
- Streamlined compensation assessments
- Delivered HIPAA/compliance-ready solution
Key Learnings
Successfully bridged AI, cloud engineering, and domain expertise (healthcare/legal) to deliver real-world solutions in sensitive, regulated industries.
Smart Building Automation Platform
Challenge
The building automation company needed a modern, scalable platform for controlling smart devices across multiple locations, with real-time data streaming from on-premise infrastructure to cloud.
Solution
Designed and delivered a comprehensive microservices-based solution:
- Mobile Applications: iOS/Android apps for device control via Bluetooth and WiFi
- Cloud Microservices: Azure-based services for authentication, access rights management, configuration, and live dashboarding
- IoT Integration: Azure IoT Hub for device connectivity and telemetry data transfer
- Infrastructure: Full CI/CD pipeline via Azure DevOps for automated deployment
Technical Stack:
Cloud: Azure (App Services, SQL Database, IoT Hub, Service Bus, Functions) | Backend: .NET microservices | Frontend: Angular web, React Native mobile | Architecture: Cloud-native, microservices, IoT-enabled | DevOps: Azure DevOps CI/CD
Team & Organization
Led and coordinated 4–7 developers, enforcing SCRUM methodology and maintaining quality standards through code reviews and automated testing.
Impact
- Delivered scalable solution handling thousands of devices
- Real-time data streaming and control from cloud
- Improved operational efficiency for building management
- Established production-grade CI/CD infrastructure
Data Masking & GDPR Compliance Solution (BizDataX)
Challenge
Enterprises needed an enterprise-grade solution for masking sensitive data (GDPR compliance) and subsetting databases for testing and analysis without exposing confidential information.
Solution
Built a comprehensive data management platform featuring:
- Multiple masking algorithms for different data types
- Support for diverse databases (SQL Server, Oracle, PostgreSQL, DB2, MarkLogic, flat files)
- Web-based UI for data discovery, schema analysis, masking configuration, and execution
- Data subsetting capabilities for safe test data generation
- Compliant with GDPR and enterprise security standards
Technical Stack:
Backend: .NET services | Frontend: ASP.NET web UI | Databases: Multi-database abstraction layer | Architecture: Modular, extensible design | DevOps: CI/CD with SonarQube and unit test coverage monitoring
Team & Leadership
Managed team of 10–14 members through full product lifecycle, implementing agile (SCRUM/Kanban) methodology with Azure DevOps for work item management.
Responsibilities included:
- Product Owner: Business requirements, feature roadmap, release planning, budget management
- Solution Architect: Defined solution architecture and database abstraction strategy
- Development Lead: Established CI/CD infrastructure, enforced code quality standards, conducted code reviews
Impact
- Deployed to multiple enterprise clients
- Successfully addressed GDPR compliance requirements
- Significantly reduced data preparation time for testing
- Maintained 95%+ code coverage and rigorous quality standards