Ahmed Zaid

debugging, documentation drafting, test scaffolding, and architectural
exploration.
• Designed and built LLM-driven workflows for domain-specific use cases, including voice-to-structured
data pipelines for medical documentation and training dataset generation for NLP fine-tuning across
52+ medical exam types.
• Experienced in evaluating AI output critically: validating for correctness, identifying hallucinations,
assessing security implications, and maintaining engineering judgment before production use.
• Exposure to prompt engineering, LLM integration patterns, multi-model orchestration (local LLMs
+ cloud models), and AI-enabled product design concepts applied in real-world engineering contexts.
• Architected and led transformation of legacy HTML portal to a scalable multi-tenant SaaS applica
tion using .NET Core and Azure cloud services, supporting 15+ concurrent customer deployments
with independent scaling per tenant.
• Designed microservices-based infrastructure using clean architecture principles, improving system
modularity, enabling independent service deployment, and reducing cross-team coupling.
• Directed database migration from SQL Server to PostgreSQL, reducing query latency by 40% and
improving horizontal scalability for high-volume medical imaging data processing.
• Built caching layer with Redis and implemented asynchronous processing patterns and Event Sourc
ing for healthcare audit compliance workflows.
• Designed and built an LLM-driven voice dictation pipeline (MatchaAI) for vascular ultrasound
documentation, including a .NET CLI tool to extract training data across 52+ exam types and
Langflow-based schemas for voice-to-JSON processing with circuit breaker resilience patterns.
• Collaborated closely with cross-functional teams (engineering, product, support) across US, and
Egypt, participating in architecture discussions and production incident resolution.
• Conducted code reviews, and contributed to engineering best practices around reliability, testing,
and release processes.

• Transformed thick-client WPF application to modern HTML5 portal using Angular and ASP.NET
Core, increasing user accessibility and reducing client-side maintenance overhead by 60%.
• Created advanced visualization tools for real-time medical imaging data, enabling 52 customer sys
tems to process exam data with sub-100ms response times.
• Led database migration from legacy SQL schema to Entity Framework ORM, improving code main
tainability and reducing query execution time by 35%.
• Managed technical escalation processes across 52 customer deployments, achieving 98% issue reso
lution rate within SLA.
• Developed 21+ custom sketch visualization tools for PACS infrastructure using SVG, C#, and WPF,
supporting complex diagnostic workflows.
• Worked in a distributed international team with regular English-language communication for tech
nical discussions, documentation, and cross-team coordination.

• Designed and implemented backend APIs for an HTML5 PACS system using React, TypeScript,
and ASP.NET Core.
• Automated database schema generation from legacy Excel documentation using Entity Framework,
reducing manual schema creation time by 75%.
• Optimized data import pipelines for medical imaging metadata, achieving 10, 000+ records/second
ingestion rate.

• Designed and deployed the Structure Sketch project, creating 21 advanced visualization tools for
real-time medical imaging interpolation within PACS infrastructure.
• Managed technical support and escalation for 52+ customer systems, maintaining 99.2% uptime
across deployments.
• Developed custom medical imaging visualization tools using C#, WPF, MySQL, and SVG libraries
supporting complex diagnostic workflows.

• Developed forecasting modules using C++ GPU threading and Python ML libraries for time series
analysis in trade systems.
• Built data transformation pipelines converting Excel-based inventory data to optimized SQL struc
tures for real-time market analysis.
• Improved system response time by 50% through algorithmic optimization and efficient data structure
utilization.