Available
Get in Touch

Case Studies

A deep dive into my technical work, ranging from enterprise database migrations to modern full-stack web applications.

AI Infrastructure & MLOps

Enterprise AI Infrastructure & Tuning

Challenge: The organization required a secure, air-gapped AI environment to analyze sensitive documents and prevent external data leakage to public models. Crucially, the solution needed to serve multiple distinct departments (Legal, IT, HR) without cross-contamination of highly specialized knowledge.

Solution: Architected and deployed a massive on-premises AI server cluster for the entire enterprise using Docker, Ollama, and OpenWebUI. I custom-tuned an open-source 116.8B parameter model into several specialized variants, providing dedicated secure models for Legal, IT, and HR.

Key Features & Impact:

  • Multimodal Vision: Integrated LLaVA and Llama Vision, allowing the AI to ingest screenshots and documents across all departments for rapid contextual analysis and troubleshooting.
  • Specialized RAG & Routing: Deployed distinct document repositories for Lawyers, IT, and HR. The AI intelligently routes queries to the correct domain-specific repository or performs web searches, drastically increasing efficiency and eliminating AI hallucinations.
  • Natural Voice AI: Implemented the tts-1 engine for fluid, human-sounding voice conversations with the assistant.
Enterprise Architecture & FinTech

Custom Financial ERP & Procurement System

Challenge: The organization needed a centralized, highly secure Enterprise Resource Planning (ERP) system to handle massive financial operations including Treasurer ledgers, paystubs, and decentralized Procurement card (p-card) transactions across all departments.

Solution: I single-handedly architected and am actively developing a comprehensive, live production ERP. The system securely ingests raw bank files, normalizes vendor data via custom rules engines, and provides sleek web interfaces for decentralized departmental data entry, splitting transactions, and reconciliation.

Key Features & Impact:

  • Massive Cost Savings & Reliability: Replaced an unreliable legacy system with a modern architecture, significantly increasing overall uptime while saving the organization an estimated $2M per year in comparable enterprise licensing fees.
  • Treasurer & Accounting Module: Accurately tracks the organization's ledgers, checks, fund transfers, and generates secure paystubs.
  • Automated Data Normalization: Built a data pipeline (functioning similarly to email rules) that ingests raw bank CSVs and automatically normalizes disparate vendor codes (e.g. AMZ018486) into clean entities (e.g. Amazon).
  • Decentralized P-Card Reconciliation: Created a secure web portal for cardholders to log in, define purchases, split transactions across multiple fund line items, and finalize historical reconciliations.
  • Immutable Auditing: Every data entry and reconciliation is securely stored in the ERP database with strict tracking of who made the change and when.
  • Admin & Fuel Tracking Dashboards: Designed high-level reporting dashboards for admins to monitor monthly transaction volumes, pending reconciliations, and specialized dashboards for fleet vehicle fuel tracking (gallons used per user/vehicle).
DevOps, Workflow Automation & Security

Enterprise IT Change Control System

Challenge: With various enterprise departments (e.g., the GIS team in the Assessor's Office) pushing updates that affect shared IT infrastructure, uncoordinated changes often led to overlapping maintenance windows, false alarms in the Security Operations Center (SOC), and unexpected downtime.

Solution: Developed a centralized, highly automated Change Control platform used by IT and expanded enterprise-wide. The system visualizes all upcoming downtime, automatically enforces approval workflows, and seamlessly integrates with monitoring tools and communication channels.

Key Features & Impact:

  • Enterprise Cost Savings: By building this custom solution internally, the organization avoids an estimated $75,000 to $100,000+ per year in comparable enterprise ITSM (IT Service Management) licensing fees.
  • Real-Time SignalR Dashboard: Visual timeline powered by SignalR WebSockets instantly pushes approved changes to the live dashboard, immediately highlighting conflicts and ensuring nobody is blindsided by maintenance.
  • Smart Departmental Routing: Reads dependency tags from our monitoring environment to automatically email specific stakeholders (e.g., Emergency Management, Assessors) whenever a system they rely on (like ARC mapping) is scheduled for downtime.
  • Automated SOC & Monitoring Integration: Automatically pauses alerts in our monitoring systems for affected servers during the change window and notifies the SOC to prevent false positive alarms.
  • Ping Messaging Integration: Integrates directly with the Ping Enterprise Messaging system (which I also built) for instant alerts on change submissions, approvals, and rejections.
  • Director Approval Workflow & Analytics: Enforces a strict review chain for directors, backed by granular graphing and metric reporting on change types, requesting departments, and rollback rates.
Real-Time Communication

Ping: Enterprise Messaging System

Challenge: The organization faced a $100,000/year quote for Slack which offered insufficient data retention (90 days vs 1 year required) and lacked critical government compliance features (FOIA auditing).

Solution: Architected and built a fully on-premise, real-time messaging platform using C# and SignalR backed by SQL Server. It provides instant communication while adhering to strict legal and accessibility standards.

Key Features & Impact:

  • $100k Annual Savings: Eliminated the need for expensive third-party SaaS licensing.
  • Government Compliance: Built-in HR audit portals for FOIA requests and harassment investigations; 1-year data retention policy enforced at the database level.
  • Full Accessibility (ADA): Engineered front-end with high-contrast modes, scalable typography, and reduced motion settings.
  • Departmental Security: Granular RBAC permissions ensure department isolation while permitting secure interdepartmental communication.
Full Stack & Automation

UserForge

Challenge: Manual creation of Active Directory accounts took 30+ minutes per user, was prone to human error, and lacked sufficient logging for security audits.

Solution: Developed a secure web-based application that standardizes and automates the entire lifecycle of user account management for the helpdesk.

Key Features & Impact:

  • 98% Efficiency Gain: Reduced account creation time from 30 minutes to 30 seconds.
  • Security & Accountability: Implemented granular permission controls and comprehensive logging for all actions.
  • Error Reduction: Standardized inputs and automated validation eliminated inconsistency and human error.
Full Stack, Workflow Automation & Security

HR Payroll & Document Management System

Challenge: The HR team spent 2 weeks/month manually calculating step-increases and generating evaluation letters. Additionally, sensitive employee documents were stored on standard network drives, causing accidental deletions, security risks, and heavy IT Helpdesk burdens.

Solution: Built a unified C# web application integrated with the ERP system. It automates payroll eligibility checks, pay adjustments, and provides a secure, self-service document portal with immutable auditing tailored specifically to HR's workflow.

Key Features & Impact:

  • Time & Cost Savings: Saved approx. 240 man-hours per month while avoiding $15k-$50k in third-party vendor licensing costs.
  • Process Automation: Automated complex 5% step-increase logic and dynamic step-raise letter generation.
  • Enhanced Security: Documents are stored in a mathematically obfuscated flat architecture to prevent unauthorized network access.
  • Immutable Auditing: Every view, download, and permission change is tracked and attributed to specific Active Directory accounts.

Note on Source Code: Due to the secure, air-gapped, and proprietary nature of these enterprise systems, source code cannot be shared publicly. I am happy to discuss system architecture, design patterns, and deployment strategies during an interview.

⇧