Portfolio & Work History
Enterprise engineering across legal tech, aviation, finance, property management, and education — production systems, not proof of concepts.
engineering
in 6 months
time reduction
at American Airlines
Witherite Law Group
Automation Engineer → Enterprise Application Architect
📄 What it does: Splits large incoming PDFs at custom markers, extracts metadata, and routes documents into Power Automate/SharePoint workflows — fully automated, high-volume, production.
🏗️ Stack: Azure Functions (Python 3.12) · Computer Vision OCR · Fuzzy string matching · Blob/Table/Queue Storage · Power Automate · SharePoint Online · Azure Monitor + OpenTelemetry
🔒 Security: SAS URL generation for time-limited, compliant document access — no permanent links, no exposure.
📊 Results:
✅ 80%+ reduction in manual handling time
✅ Processes hundreds of documents/day, reliable up to 400+ page files
✅ Automated error escalation minimized workflow interruptions
🔧 Azure Functions · Blob/Table/Queue Storage · Computer Vision OCR · PyMuPDF · Levenshtein · Power Automate · SharePoint Online · Azure Monitor · Microsoft Visio
⚖️ What it does: HTTP-triggered Azure Function queries SQL for legal case data, fuzzy-matches witnesses on DLN, Plate, VIN, DOB, SSN, deduplicates conflicts, auto-generates PDF reports → Blob + Table Storage. Hundreds of records per run.
🏗️ Flow: HTTP trigger → SQL query → rapidfuzz matching → deduplication → reportlab PDF → Blob Storage + Table audit log
📊 Results:
✅ Eliminated manual conflict review — fully automated end to end
✅ Hundreds of records/run, proven on large datasets
✅ Secure storage + access controls protect sensitive legal data
✅ Comprehensive test suite validates deduplication logic
🔧 Azure Functions · Blob/Table Storage · SQL Database · rapidfuzz · PyPDF2 · reportlab · Azure SDKs
👤 What it does: Fully automates new-hire onboarding from ADP → Microsoft Entra ID. User creation, group assignments (dept, payroll, location), manager relationships — zero manual HR steps.
🏗️ Flow: ADP HR API (paginated) → JSON schema mapping engine → Microsoft Graph API → Entra ID user + groups → mail-enabled groups via Azure Automation Runbook webhook
🔒 Security: Certificate-based auth · Managed Identity · time-limited credentials · no secrets in code
📊 Results:
✅ Onboarding reduced from manual effort to fully automated
✅ Handles large hire batches with extensible JSON/env-var mapping logic
✅ Downstream M365 processes accelerated from day one
🔧 Azure Functions (Python) · Microsoft Graph API · ADP HR API · Azure Automation Runbook · Blob/Table Storage · JSON Schema · OpenTelemetry · Azure Monitor
🗄️ What it does: Ingests multi-source raw CSVs into a centralized Azure Data Warehouse — dynamically infers schema, syncs SQL columns, deduplicates against unique keys, converts to Parquet → live Power BI dashboards.
🏗️ Flow: Raw CSV (Blob) → Pandas schema inference → SQL column sync → unique key deduplication → PyArrow Parquet → Power BI
📋 Governance: Azure Table Storage tracks metadata, audit logs, and dynamic folder mappings for full data lineage.
📊 Results:
✅ Manual ETL effort eliminated — fully automated onboarding
✅ Consistent, accurate reporting via automated schema management
✅ Real-time Power BI analytics from day of ingestion
🔧 Azure Functions (Python) · Blob/Table Storage · SQL Database · Pandas · PyArrow · Power BI · Azure SDKs
🛡️ As lead engineer, I set and enforced error handling + logging standards adopted across all projects team-wide.
✅ Input validation + structured error responses on every Azure Function
✅ try/except on all external calls — full stack traces logged
✅ Azure Monitor + Application Insights — real-time diagnostics & distributed tracing
✅ Retry logic + graceful degradation — only critical failures halt processing
✅ Centralized structured logging — real-time and historical analysis
✅ Auto help desk ticket creation on persistent/critical failures
✅ Admin audit logging table for compliance & process tracking
✅ Patterns enforced via code reviews + team training
📈 Result: Reduced downtime, faster incident response, measurably improved system reliability across the org.
🏗️ Led design & implementation of the Halo Multi-Tenant SaaS Platform — enterprise-grade legal CRM built on Azure, from scratch, as sole architect.
⚙️ Architecture highlights:
✅ Module-based CRM — pluggable tenant data stores + dynamic UI composition
✅ Multi-tenant data isolation + failure domain separation + region-aware provisioning
✅ Immutable append-only audit logging on all critical actions
✅ RBAC + tenant-scoped secret management + control/data plane separation
✅ Dead-letter queues, retry policies, configurable log retention
✅ Centralized monitoring, distributed tracing, real-time alerting
✅ CI/CD, disaster recovery, cost optimization across Dev/UAT/Prod
✅ SLA/RTO/RPO documentation authored and maintained
📈 Result: Halo became the reference architecture for enterprise SaaS reliability at the org — faster feature delivery, higher quality, lower operational risk.
🧾 What it does: End-to-end accounts payable automation — invoices are OCR-scanned via Azure Computer Vision, validated against CRM data for coding accuracy and amount confirmation, then delivered directly into Acumatica via REST API. Zero manual invoice entry. Zero coding errors.
🏗️ Flow: Invoice (PDF/image) → Azure Computer Vision OCR → extracted fields → CRM integration (validation + coding confirmation) → Acumatica REST API (payable delivery) → Power Apps dashboard (real-time flow status)
🔒 Security: Managed Identity for Acumatica API auth · SAS URL access for document storage · structured audit logging on every invoice state transition
📊 Results:
✅ Manual invoice coding eliminated — fully automated AP entry into Acumatica
✅ CRM validation catches coding mismatches before they hit the ledger
✅ Power Apps dashboard gives finance team real-time visibility across all invoice states
✅ Scales to high-volume invoice batches without additional headcount
🔧 Azure Functions (Python) · Azure Computer Vision OCR · Acumatica REST API · CRM Integration · Power Apps · Blob/Table Storage · Managed Identity · Azure Monitor
🎯 6 production systems. 6 months. One engineer promoted to architect.
From mailroom automation to enterprise SaaS architecture — I delivered technical vision, operational discipline, and team-wide engineering standards. The Halo platform and the error handling culture I built are both still running and growing today.
6 production systems in 6 months — OCR, AP/AR automation, AI agents, and multi-tenant SaaS architecture powering the Halo platform.
American Airlines
Engineer / Developer / Project Manager
Designed Dev/Stage/QA/Prod Azure environments for enterprise employee-facing apps. CI/CD via Azure DevOps, Jenkins, Python — 30% faster deployment cycles. Docker + Kubernetes standardization across all stages — 40% fewer deployment errors, zero-downtime releases.
ETL pipelines with Azure Data Factory + Databricks: real-time baggage data for 20,000+ users globally, 25% lower latency. Kafka + Azure Data Lake + HDFS for streaming ingestion under corporate governance. Azure ML predictive analytics for resource demand forecasting.
Automated vulnerability detection across 60+ repositories — Coverity, Black Duck, Nucleus — 90% reduction in manual checks. Python/Bash scripts for ETL orchestration, resource scaling, and deployment automation.
Terraform + Azure Resource Manager — deployment time cut by 82.98%. High-availability DevOps architecture supporting real-time staff management systems.
Aligned 7 teams (~60 members each) on security protocols, deployment strategy, and DevOps best practices. Introduced CI/CD onboarding docs — accelerated adoption team-wide.
🔧 See the full tech stack in the Tech Stack accordion below.
🛬 FAA — Flight Real-Time Data Mapping
Integrated + mapped real-time FAA flight data into gate agent internal apps. Bi-weekly stakeholder sessions to resolve accuracy and mapping issues. → See the Industry Engineering accordion below.
🧳 Brock Solutions — Baggage Carousel Visualization
Connected AA baggage scan/location data to Brock's visualization systems — second-by-second updates for airport staff globally. Built + maintained secure APIs. Weekly alignment on features and config. → See the Industry Engineering accordion below.
🌐 Siemens — Networking Real-Time Feeds
Collaborated on port upgrades, firewall enhancements, NAT/PAT reconfigurations. Led end-to-end testing of real-time data feeds under high traffic. Weekly health checks and upgrade coordination. → See the Industry Engineering accordion below.
Data Engineering: Azure Data Lake, Databricks, Cloud Storage
ML / Automation / Visualization: Python, Power Automate, Power BI
Security: Coverity, Black Duck, Tanium, Nucleus
✈️ Baggage Handling System
- Baggage Handlers: Loading, unloading, and transferring baggage.
- Airport Staff: Customer service and security roles supporting airport operations.
- Passengers: Directly affected by baggage delays or losses.
- Airline Operations: Real-time information for efficient flight operations and customer service.
- Purpose: Track and manage baggage movement network-wide, ensuring efficient operations and reducing losses.
- Resolve: Address baggage delays, lost luggage, and handling inefficiencies.
- Predictive analytics: ML to forecast baggage trends and flag potential issues.
- Automation: Automate routine tasks to improve efficiency and reduce errors.
- Data integration: Connect flight information with passenger and baggage data systems.
- Data Engineering: Azure Data Lake, Databricks, Cloud Storage
- ML / Automation / Visualization: Python, Power Automate, Power BI
- Security: Coverity, Black Duck, Tanium, Nucleus
🛂 In-Airport Staff Management
- Gate Agents: Manage passenger boarding and gate operations.
- Pilots and Flight Crew: Access schedules and flight assignments.
- Airport Staff: Customer service, maintenance, and security roles.
- Purpose: Centralized platform for in-airport staff management ensuring accurate tracking of flight information.
- Resolve: Eliminate manual data entry, information access delays, and inefficiencies in staff coordination.
- Data analytics for operational trend identification and optimization.
- Mobile accessibility for on-the-go staff updates.
- Staff scheduling: track and manage shifts and assignments.
- Real-time flight status, gate assignments, and boarding information.
- Terminal and gate management with resource allocation.
- Frontend/Backend: React, Angular, or Vue.js · Node.js or Python (Django/Flask)
- Database: PostgreSQL or MongoDB
- Cloud/DevOps: Azure, Git, Jenkins/CircleCI, Docker/Kubernetes
Citibank
DevOps / Data Engineer
ETL optimization with Azure Databricks, PySpark, and Java — 30% faster ingestion across large-scale financial datasets. Azure Data Lake + Hadoop (HDFS, Hive, Kafka) + SQL for high-performance transformation under compliance requirements. Azure-based pipelines supporting in-house API integrations for end-user analytics.
CI/CD with Azure DevOps, Jenkins, Kubernetes — consistent delivery across Big Data apps. Terraform + ARM for automated environment provisioning. Maven, SonarQube, GitBash as automated code quality gates — 25% fewer deployment errors.
Java analytics apps deployed via Docker + Kubernetes — scalable, environment-consistent. Python/Bash scripts for recurring ETL tasks and monitoring — 20% reduction in processing time.
Secure, compliant financial data processing at scale. Agile delivery across multi-functional teams for mission-critical analytics. → See the Tech Stack accordion below.
Data Engineering: Python, PySpark, HDFS, HBase, Hive, Hadoop Ecosystem
ETL: Databricks, Python, PySpark
DevOps: Jenkins, BMC RLM, Autosys, Docker, Kubernetes, CI/CD, GitBash, Java/J2EE
💰 Data Platform & Analytics
- Data Scientists: Utilize developed tools and platforms for analysis and modeling.
- Business Analysts: Rely on data insights for informed decisions.
- IT Teams: Maintain and support the data infrastructure.
- Data ingestion: Extract data from various sources and load into the platform.
- Data transformation: Clean, transform, and prepare data for analysis.
- Data governance: Implement practices to ensure data quality and consistency.
- Data Engineering: Python, PySpark, HDFS, HBase, Hive, Hadoop Ecosystem
- ETL: Databricks, Python, PySpark
- DevOps: Jenkins, BMC RLM, Autosys, Docker, Kubernetes, CI/CD, GitBash, Java/J2EE
Barvin
Property Manager
Excel-based operational reports tracking occupancy, maintenance, and financial performance. SQL used to extract and analyze budget variances — data-driven decisions, not gut feel.
Scripted workflows for rent reminders and maintenance tracking — improved response times, reduced manual follow-up. Standardized data integration across PM systems into central databases for consistent insights.
Tenant communications managed via integrated systems — high satisfaction, prompt resolution.
Code Ninjas
DevOps Director
Streamlined cross-team workflows using Azure Boards — stakeholder alignment, improved timelines, client feedback loops integrated directly into delivery cycles.
Used analytics to manage and grow $300k+ in data services — refined acquisition strategies to expand market reach.
🔧 See the full tech stack in the Tech Stack accordion below.
Homebase for workforce scheduling and Vonage for VoIP communication systems — streamlining operations and client communication for the studio.