Fatima Macias

Evaluating and refining AI model outputs for quality, safety, and alignment as an independent contractor.

• Evaluate AI-generated responses for quality, accuracy, safety, and alignment across diverse task domains.
• Provide structured feedback and ratings used to fine-tune and improve large language models.
• Operate under strict confidentiality agreements; specific clients, projects, and deliverables are proprietary and not publicly disclosed.

Architecting multi-provider AI systems and accelerating delivery with AI-assisted development workflows.

• Architected a configurable AI decision-making system supporting multiple LLM providers (Claude, OpenAI, Gemini) with built-in cost management, enabling dynamic provider switching based on task requirements.
• Designed and implemented provider-agnostic AI pipelines, abstracting model-specific logic to ensure flexibility and cost efficiency across production AI workflows.
• Adopted spec-driven development, writing detailed implementation specs and using TDD to translate AI-assisted plans into reliable, well-tested code.
• Accelerated development velocity by integrating AI coding assistants (Claude, GitHub Copilot, Sonar) into daily workflows, improving code quality and reducing implementation time.
• Built an AI-powered study platform using Python, FastAPI, and Gemini API to auto-generate structured educational content from video sources.

Led DDD-based insurance platforms, cutting API response times 55% and reaching 98% test coverage.

• Architected self-service insurance comparison platforms (auto, health, life) using Domain-Driven Design (DDD) on ABP Framework, modeling complex broker and policy domains into clean, bounded contexts with built-in security layers and audit logging.
• Leveraged ABP Framework's built-in modules for authorization, multi-tenancy, and logging, reducing cross-cutting concerns and accelerating delivery of core business features.
• Integrated 3 payment gateways across insurance platforms, expanding customer payment options and increasing processing flexibility by 40%.
• Achieved 55% reduction in API response times through SQL query optimization, strategic indexing, and distributed caching.
• Reached 98% test coverage with xUnit, cutting post-deployment bugs by 90%, and implemented CI/CD pipelines in Azure DevOps for reliable software delivery.

Automated financial operations and ETL pipelines, boosting productivity 25% across finance teams.

• Engineered a centralized payment management system integrating 7 financial institutions, automating 100% of manual supplier payments and boosting department productivity by 25%.
• Built an ETL pipeline integrated with JD Edwards ERP, automating end-to-end financial data extraction and reducing the finance team's workload by 25%.
• Modernized and modularized legacy codebases, reducing operating costs by 25% and improving system scalability.
• Engineered SQL Server solutions including schema design, stored procedures, and query optimization for high-volume data workloads.
• Developed an electronic invoicing module using ASP.NET MVC and SQL Server, ensuring full regulatory compliance.

Built port operations and billing systems integrating weighing hardware with invoicing software.

• Engineered a port operations system integrating truck scale/weighing hardware with billing software, automatically calculating storage, inspection, and port fees for bulk and weight-based cargo.
• Developed electronic invoicing and shift management systems, improving operational traceability and compliance across port workflows.
• Collaborated with port operations teams and analysts to translate complex domain requirements into production-grade software solutions.

Automated financial reporting and modernized legacy VB6 systems to VB.NET.

• Automated financial report generation, boosting accounting department productivity by 25% through seamless data integration.
• Modernized VB6 legacy code to VB.NET, improving compatibility, maintainability, and alignment with modern development standards.