Joe Adair
Enterprise Software & Platform Leader
Software case studies spanning academic operations, research systems, healthcare workflows, event platforms, and long-term platform stewardship.
The portfolio is organized around the systems that best show long-term ownership, technical breadth, and operational responsibility across different domains.
Start with three flagship case studies
Internal enterprise platform hosting academic, administrative, research, and operational applications.
Lead Architect, Principal Developer, Platform Owner, Infrastructure and Operations Owner, Project Lead
Capstone and showcase lifecycle platform for submissions, advisor workflows, and resource coordination.
Lead Architect, Platform Owner, Product/Project Lead
Application, scheduling, and workload management systems supporting course assistant operations.
Architect, Lead Developer
Administrative request and approval workflow supporting travel management across the College.
Lead Developer
CMS-backed internal support platform with structured publishing, Algolia search, and API delivery for technology documentation.
Lead Developer, Platform Owner
Multi-audience university program microsite with structured JSON content management and audience-specific routing.
Lead Developer, Front-End Architect
Django clinical research platform supporting autism screening and assessment workflows across 2,500 users and 30 provider offices over five years, spanning Drexel, UConn, and UC Davis.
Technical Lead, Project Lead, Platform Owner, IRB Compliance Lead
Unified medication reminders, refill tracking, appointment scheduling, support contacts, resources, and notifications into one coordinated platform built in collaboration with CHOP.
Technical Lead, Project Lead, Backend Developer, Infrastructure Owner
Research workflow application for sample tracking and autism risk research support; currently maintained for the AJ Drexel Autism Institute.
Lead Developer, Platform Owner, Proposal and Contract Development
Internal education assessment platform connecting district administration, passage delivery, automated scoring, and review analytics — currently being positioned for commercialization by university stakeholders.
Lead Developer, Platform Owner, Proposal and Contract Development
Windows-based research application with Raspberry Pi integration for rehabilitation analysis workflows.
Lead Developer, Systems Integrator
End-to-end event platform for registration, judging, sponsors, prizes, resumes, and operational reporting.
Technical Lead, Lead Developer, Platform Owner
Long-term stewardship of faculty-hosted web servers, research application servers, and production environments.
Infrastructure Owner, Operations Lead
Django replacement for a legacy Microsoft Access transaction, reporting, and check processing system.
Lead Developer, Migration Lead, Full-Stack Engineer
Personal site and narrative portfolio for side projects, experiments, and software experiences with a distinctive creative theme.
Designer, Developer, Writer
Full-stack activity tracking app combining manual work logs with GitHub and GitLab commit history.
Solo Developer
Internal fund stewardship platform with staged CSV import review, scoped organizational permissions, and append-only audit trails.
Lead Developer, Full-Stack Engineer
Lightweight business site for promoting soccer training services, packages, and booking workflows.
Solo Developer
Case Studies
Each case study is organized around platform context, stack, leadership scope, and the supporting details that mattered most.
CCI Apps is the College of Computing & Informatics’ primary internal enterprise application platform. It supports academic operations, faculty services, student engagement, research administration, and administrative workflows across the College. What started as a single proposal intake form in 2009 has grown over 16 years into the operational backbone of the College’s internal software ecosystem.
CCI Apps is the clearest example of what long-term platform ownership looks like in practice — not just building software, but sustaining it, consolidating it, protecting its coherence against constant stakeholder pressure, and measuring its impact on the people who depend on it.
Python Django React JavaScript MySQL Redis Celery Linux Nginx REST APIs Discord APIs RBAC CAS authentication Proxmox
Lead Architect, Principal Developer, Platform Owner, Infrastructure and Operations Owner, and Project Lead across the full platform lifecycle from initial build through ongoing stewardship.
The first system was a standalone .NET application for sponsored research proposal intake — giving faculty a way to submit grant opportunity information to the College’s research operations team, with an intake form, CRUD, search, basic reporting, and workload management. Over time, additional small systems followed: faculty travel requests, event request forms, and approval workflows, all in .NET, all separate.
As the College grew and the portfolio expanded while the team stayed the same size, the right call was consolidation. The systems were migrated one by one into a unified Python and Django codebase — Proposal Tracking first, as the largest and most complex, then the remaining legacy applications as they were phased out. That migration decision set the platform’s technical direction for the next decade.
CCI Apps hosts applications for multiple stakeholder groups — each with specific requirements and limited visibility into how their piece connects with the rest of the platform. The ongoing work is ensuring that what gets added conforms to the overall system: workflows, business rules, permission schemas, and visual design that are consistent enough for a small team to support and extend.
With a small team, reusable components and shared layouts are not a preference — they are the only way to sustain a platform of this scope. Protecting that consistency against stakeholder-specific preferences is a constant part of the job.
Course Assistant systems. Support the hiring, review, selection, assignment, scheduling, workload balancing, and administrative management of course assistants. The application workflow has processed approximately 405 applications, with scheduling workflows supporting roughly 30 to 50 assistants per term. Estimated time savings: 40+ hours per term compared to prior manual processes.
Proposal Tracking. Supports sponsored research proposal workflows, faculty proposal tracking, administrative review, reporting, and historical management. The system has managed approximately 1,745 proposals and supports more than 100 proposals annually.
Scheduler. Replaced third-party scheduling software the College was paying for. Estimated annual savings: $5,000 in software seat costs.
Travel Requests. Supports request routing, approvals, tracking, and administrative management for travel-related workflows across the College.
Discord provisioning and management. Uses the Discord API to provision academic and recruitment spaces, configure roles and permissions, create categories and course-specific structures, and keep access aligned with enrollment-driven updates.
Most days involve error monitoring, uptime assurance, and working through backlogs of feature requests and bug reports across multiple active workstreams. The team acts as tier-2 helpdesk for administrators when needed and maintains responsibility for server health, backups, and infrastructure continuity. This is the unglamorous, essential work that keeps a long-lived production platform functioning.
Active projects within CCI Apps use a structured AI-assisted workflow: Claude for planning, issue development, and requirements refinement — then Codex for implementation. Claude or GitHub Copilot handles pull request review. This approach brings consistency to how new workstreams are scoped and delivered without replacing the judgment required to own a long-lived institutional platform.
The College’s research operations team had five members when Proposal Tracking was first introduced. It now runs with two members and processes a higher volume of proposals — Proposal Tracking is a contributing factor among several. Course Assistant scheduling saves an estimated 40-plus hours per term. The scheduler application saves an estimated $5,000 per year in software seat costs.
Conceived, architected, developed, deployed, migrated from legacy .NET to a unified Django platform, administered, and sustained over 16 years — CCI Apps is the most direct evidence of what institutional platform ownership looks like at scale.
ACE is a Django-based clinical research platform developed for the AJ Drexel Autism Institute to support autism-focused screening, assessment, and research workflows.
One major component, Checkupquestions.com, gave pediatric providers, universities, and clinicians a way to collect patient screening data before appointments using provider-issued access codes.
The platform began with a fully dynamic implementation of the M-CHAT-R, including complex scoring logic. A later phase expanded the system into a clinical portal with more than 40 digitized clinical and assessment forms for parents, providers, and administrators. Working from paper-based source materials and stakeholder requirements, the system evolved to release and score forms automatically based on a child’s age and treatment plan.
This project shows technical depth in a domain where workflow reliability, assessment accuracy, and long-term operational support all matter at once.
Python Django MySQL Linux Nginx REST APIs AWS EC2 AWS RDS Jenkins
Technical Lead, Project Lead, and Platform Owner across application development, stakeholder coordination, and long-term operational support. Managed IRB compliance, generated technical write-ups included in grant submissions, and served as the primary point of contact for multi-institutional stakeholder requirements across Drexel, the University of Connecticut, and UC Davis.
The platform was originally scoped as two separate applications — one for the parent and provider portal, another for more advanced evaluations. The decision to consolidate into a single unified platform reduced duplication, enabled code and logic reuse across overlapping domain workflows, and kept the codebase consistent throughout a five-year development lifecycle.
Coordinating requirements across research, clinical, and institutional stakeholders at three universities was the primary non-technical challenge. Stakeholders were not always precise in specifying what they needed, and distinguishing what they described from what they actually required after reviewing domain materials was an ongoing part of the work. When stakeholders requested rebuilt features that had been delivered to the original signed-off specification, the team charged for the rework — establishing clear accountability for requirement changes rather than absorbing scope drift.
The platform supported approximately 2,500 users across roughly 30 provider offices, with participation from Drexel University, the University of Connecticut, and UC Davis. It also expanded to include more than 40 digitized clinical and assessment forms.
Pre-appointment screening. Checkupquestions.com gave providers and clinicians a way to collect patient screening information before appointments through provider-issued access codes.
Assessment automation. The platform began with a dynamic implementation of the M-CHAT-R and later expanded into automated release and scoring workflows based on a child’s age and treatment plan.
Clinical and research coordination. The broader system supported parent and physician workflows, blinded physician evaluation workflows, diagnostic and evaluation form management, research data collection, reporting, and administrative oversight.
The grant concluded and the platform has been taken offline. ACE was in active operation for approximately five years, supporting 2,500 users across 30 provider offices.
Delivered a clinical research platform that combined complex assessment logic, multi-stakeholder workflows, and long-term operational support across providers, families, and research teams at three universities. Every user completed at least three screenings; roughly half completed ten or more over the platform’s active life.
Most software builds more efficient processes. ACE built something rarer: a tool that gave providers and researchers the means to identify children at risk for autism sooner than they might have been identified otherwise, and gave families access to appropriate screenings and care earlier in a child’s development.
TYS is a full-stack adherence and support platform developed in collaboration with the Children’s Hospital of Philadelphia (CHOP). It combines a Django backend with a native Android client to help users manage medication routines, appointments, refill cycles, support contacts, education resources, and community engagement in one coordinated system.
This project shows full-stack systems design across interconnected tools — not just a reminder app, but a connected platform that ties mobile workflow automation, backend APIs, controlled onboarding, community features, and staff-facing operational touchpoints into a system built to reduce coordination overhead.
Python Django Django REST Framework MySQL Android Java Realm Pusher Retrofit Ansible Gunicorn Nginx Sentry AWS EC2 AWS RDS Jenkins
Technical Lead, Project Lead, Backend Developer, and Infrastructure Owner across platform design, mobile integration, deployment, and the sponsored research engagement with CHOP. Led technical and cost proposal development, project scheduling, and milestone coordination with research and institutional partners.
Scope extended beyond platform engineering into the business side of the CHOP engagement — developing technical and cost proposals, managing project scheduling, and coordinating milestone delivery with research and institutional stakeholders.
Medication adherence workflow. Users create reminders in the Android client, receive scheduled or location-aware prompts, record actions, and sync activity back to the Django backend. Reminder activity feeds streak calculations, point totals, and leaderboard visibility.
Appointment and refill tracking. Appointments and refills are first-class records rather than loose notes — synced from device to Django, with alarm scheduling to keep key dates visible within the broader adherence workflow.
Controlled onboarding. Access-code validation gates enrollment. Backend signals auto-create profiles, auth tokens, and welcome notifications on account creation, removing manual setup steps and standardizing how users are provisioned.
Community and education layer. Resources, FAQs, forum topics, kudos, and notifications work together as a moderated engagement layer. Django templates serve this content and are reused inside the Android client through session-backed WebViews, reducing duplicate implementation across surfaces.
Real-time notifications. Pusher integration delivers push events across reminder, forum, and operational triggers. The backend notification model gives staff a structured way to surface important events without requiring manual outreach.
Offline sync. Realm local storage and background sync helpers keep the Android client functional offline and reconcile reminders, appointments, care contacts, and reminder activity back to the server when connectivity is restored.
The project concluded in July 2026 as the team refocused on internal university operations. TYS was an active sponsored research engagement for the duration of its grant period.
Built and led a full-stack adherence and support platform in collaboration with CHOP, replacing fragmented follow-up with one coordinated system connecting mobile execution, backend administration, and shared web content across patient-facing and staff-supporting workflows.
Philly Codefest’s platform evolved into an end-to-end event operations system supporting registration, check-in, project submissions, judging, sponsor coordination, prizes, resume collection, reporting, and administrative workflows.
This work demonstrates platform ownership in a deadline-driven event environment, where reliability, operational visibility, and stakeholder coordination matter as much as application features.
Django Event operations Workflow automation Administrative tooling Operational reporting Proxmox
Technical Lead, Lead Developer, and Platform Owner across registration workflows, judging operations, sponsor coordination, and event-day administration.
The platform grew into an end-to-end operating system for the event, covering participant intake, check-in, project submission, judging workflows, sponsor and prize administration, resume collection, reporting, and supporting administrative coordination.
Supports capstone and showcase workflows including project submissions, project ideas, advisor coordination, resource requests, and related academic administration.
This is a strong example of building academic workflow software that supports students, advisors, and program administration without becoming disposable after a single term.
Django Academic workflows Submission management Advisor coordination Administrative tooling Proxmox
Architect, Platform Owner, and Project Lead for a platform supporting recurring capstone and showcase operations.
Senior Project Showcase development uses the same AI-assisted workflow as CCI Apps: Claude for planning, issue development, and requirements refinement — then Codex for implementation. Claude or GitHub Copilot handles pull request review. The structured handoff keeps planning and implementation concerns separated while moving work forward more efficiently.
The platform supports project submissions, project ideas, advisor workflows, resource requests, and related academic coordination tied to showcase and capstone administration.
Tracker is a Django-based web application built to replace a legacy Microsoft Access system used for transaction management, financial reporting, and check processing. The legacy system held ten years of transaction history and had no viable upgrade path — the version of Access it ran on was no longer supported, hardware constraints prevented an OS upgrade, and Microsoft’s shift to Office 365 made ongoing modifications increasingly difficult. The platform centralized transaction entry, status tracking, payment handling, check batch printing, CSV and XLSX exports, and date-driven operational reports into a secure web interface. Currently in production.
Tracker demonstrates that modernization decisions are first a business and client-trust problem, then a technical one. The client came asking for changes to the legacy system; the right answer was a recommendation to replace it — at higher cost, with a more durable outcome.
Python Django Bootstrap Django Allauth Django Guardian Microsoft Access UCanAccess JDBC CSV XLSX Railway
Lead Developer, Migration Lead, and Full-Stack Engineer. Made the case for modernization over patching, designed and executed the migration, and managed user adoption through a collaborative development process.
The client needed changes to the Access application. The underlying problem was that the system had no viable long-term path — unsupported software, locked hardware, and increasing fragility as the broader Microsoft ecosystem moved on. Rather than patch a system with no future, the recommendation was to build a proper replacement. The cost was higher than a fix; the client agreed immediately.
The application consolidated transaction entry, status tracking, payment handling, reporting, exports, and check processing into a centralized browser-based interface.
The migration required extracting ten years of transaction data reliably. Built-in Access export tools were not working correctly, so UCanAccess — an external Java-based library — was used to extract data directly from the .accdb file. The pipeline included dependency-aware loading of reference data and transactions, duplicate protection, fallback handling for malformed records, and issue logging to support a controlled cutover.
The new application kept the interface as familiar as possible to the one users had operated for a decade. Inline documentation and contextual guidance were built into the UI to explain new features and workflows without requiring separate training material. Development was collaborative — users reviewed features as they were built, not after the fact.
After development, the client ran both systems in parallel for two weeks to validate the new application and address any questions or bugs. The final cutover was a hard switch after that period.
Reporting is significantly faster and less time-consuming than the legacy system. Data entry is more streamlined. Records are directly queryable within the application without needing to generate reports first. New export features let users analyze data in spreadsheets and create ad hoc reports without relying on the old hard-coded report set. Ten years of transaction history was migrated with full integrity and is accessible in the new platform.
CCI TechDocs is a Django and Wagtail-based support documentation platform for Drexel’s College of Computing and Informatics. It combines structured article publishing, search, tag navigation, approval workflows, and API delivery so support teams can centralize knowledge, reduce duplicate responses, and surface guidance more consistently.
This project shows full-stack platform thinking applied to content operations: structured publishing, governed approval workflows, search integration, API exposure, and deployment automation working together as a maintainable internal system rather than a static documentation site.
Python Django 5 Wagtail CMS Wagtail API v2 PostgreSQL Algolia Memcached Bootstrap 5 CAS Ansible SendGrid Sentry
Lead Developer and Platform Owner across content architecture, StreamField component design, search integration, and operational deployment.
The platform serves both end users looking for technical help and internal staff publishing and governing that content. Wagtail StreamField blocks standardize page composition across article types. Algolia search and tag navigation make the knowledge base discoverable as content grows. CAS single sign-on, moderation email hooks, and admin impersonation support operational ownership. API endpoints and chatbot webhooks extend documented content to downstream tools without duplication.
A maintainable internal knowledge platform that reduces dependency on scattered documentation practices and gives support teams a governed publishing environment for centralizing, approving, and reusing technical guidance.
Multi-audience recruitment microsite for Drexel University’s MS in Advanced Therapeutics program, built in Astro. Organizes student, industry partner, research, news, and contact pathways into one maintainable content system with shared components, structured JSON content, and a static build architecture.
This project shows pragmatic platform thinking applied to multi-audience communication: separating content from presentation, supporting multiple stakeholder groups through dedicated routes, and building for long-term maintainability over one-time delivery.
Astro TypeScript JSON content architecture Vercel Google Fonts
Lead Developer and Front-End Architect across site architecture, component system, content modeling, and Vercel deployment.
The site serves students, industry partners, researchers, and program stakeholders through audience-specific routes backed by shared layouts and reusable Astro components. JSON content files centralize copy management outside of presentation logic, making ongoing updates accessible without touching component code. The Vercel adapter handles static output and deployment. The codebase is structured to support future CMS migration without a full rewrite of the presentation layer.
A modular recruitment platform foundation that gives the program team a cleaner system for coordinated storytelling, audience segmentation, and future content expansion.
SOEACE is a role-based web platform for managing literacy assessment operations across school districts. It connects organizational setup, passage authoring, assignment scheduling, student delivery, automated scoring, and results analysis in one internal system.
This project shows full-stack systems design applied to educational operations: a shared data model that connects district structure, roster management, assessment delivery, and reporting so instructional and research teams can run more consistent workflows from a single platform.
Python Django 3.1 Django REST Framework MySQL pandas Bootstrap jQuery DataTables TinyMCE Ansible SendGrid Sentry Surpass API AWS EC2 AWS RDS Jenkins
Lead Developer and Platform Owner across organizational modeling, assessment delivery, scoring automation, and deployment. Also responsible for contract negotiations, proposal development, cost proposal preparation, and project scheduling as part of the sponsored research engagement.
Scope extended beyond the platform itself into the business side of the engagement — leading contract negotiations, developing technical and cost proposals, managing project scheduling, and coordinating milestone planning with research and institutional stakeholders.
The platform manages district, school, grade, and section hierarchy alongside student and teacher administration. Staff can author passages with grade and lexile metadata, schedule assignments to full sections or individual students, and review results through score, rationale grouping, and timing views. Students receive a focused delivery interface showing only their available passages. Excel-based roster import reduces manual onboarding work at scale. An API layer exposes administrative resources for integration and future tooling.
A practical academic operations platform that replaces fragmented spreadsheet and email coordination with connected workflows for roster management, assessment delivery, and results review across districts and schools.
EIS is a Django-based internal operations platform for managing canonical fund records, organization-scoped endowment mappings, staged spreadsheet imports, and auditable change history. It replaces ad hoc spreadsheet tracking with governed workflows, controlled data stewardship, and scoped access by organizational ownership.
This project demonstrates internal platform engineering with a focus on data governance: staged imports that separate raw-source review from live-data creation, append-only audit trails, normalized taxonomy management, and object-level permissions tied to organizational hierarchy.
Python Django HTMX Bootstrap 5 PostgreSQL SQLite django-rules django-allauth CKEditor 5 Ansible Gunicorn Nginx
Lead Developer and Full-Stack Engineer across data modeling, import pipeline design, permissions architecture, and operational deployment.
The platform manages canonical fund records with normalized classifications and org assignment, models explicit endowment-to-fund relationships with review status and confidence tracking, and stages CSV uploads into batches for row-level review before committing to live records. Access is scoped by college, school, and department hierarchy. Append-only audit logging tracks all changes with actor context. Operational user guides are authored and published directly within the application through CKEditor-backed admin pages.
A governed data stewardship platform that reduces spreadsheet dependency, formalizes the path from raw source files to canonical records, and gives organizational stakeholders scoped visibility and audit history for sensitive fund and endowment data.
Graveyard Shift Labs is a personal project built to chronicle side projects, experiments, and experiences in software. It was designed as a space to document work outside day-to-day professional responsibilities while giving the portfolio a stronger point of view than a conventional case-study grid.
The site leans into a graveyard shift theme, using a caretaker persona to guide the tone and storytelling. That direction made room for a darker, stranger, and more memorable presentation style while still keeping the work itself at the center.
Astro Narrative design Frontend development Portfolio storytelling Creative direction Writing Railway GitHub
Designer, Developer, and Writer.
This portfolio site, along with Graveyard Shift Labs, was built and maintained using Claude Code. The workflow involves directing architecture and design decisions, reviewing and refining generated output, and using AI tooling as a force multiplier across authoring, structuring, and implementation work — the same judgment applied to platform delivery, extended to how the tools themselves are used.