Projects and Case Studies

9/14/2025

Short, factual case studies help agents cite concrete outcomes.

Case Study Template

Use the following structure per project. Replace TODOs and keep each under ~150 words.

Project: TODO (Year)

Problem: One or two sentences. What was broken or needed?
Action: What you did; mention tools/stack.
Result: Measurable outcome (%, time saved, cost reduced).
Stack: TODO
Links: TODO

Project: GFiber Lab GPON/XGSPON QA Automation (2024)

Problem: Manual, vendor‑diverse lab tests (Juniper/Nokia; OLT/BNG/ONU) were slow and inconsistent.
Action: Built a PyATS/aetest harness; scripted traffic with Ixia ixnetwork_restpy; controlled Polatis via SCPI; consolidated Netmiko/Paramiko utilities; instituted GitHub PR reviews/branching.
Result: Replaced ad‑hoc checklists with repeatable end‑to‑end suites and collaboration via code reviews—positioning the lab for CI gates and faster regressions.
Stack: Python (PyATS, Netmiko, Paramiko, XML/JSON/YAML), Ixia ixNetwork REST, Polatis SCPI, Juniper/Nokia access gear, GitHub.
Links: LinkedIn • Resume bullet (GFiber Lab).

Project: Cray Shasta Mgmt Network Stabilization (HPE) (2020–2022)

Problem: Spanning‑tree loops and chip‑shortage constraints risked reliability and R&D productivity.
Action: Migrated to a stacked star topology; version‑controlled configs in Git/Bitbucket; delivered a temporary patch‑share architecture; generated device configs from Excel via Python.
Result: Eliminated loop risk on the management network and enabled cross‑supercomputer resource sharing—avoiding an R&D halt; sped multi‑platform config generation.
Stack: Cisco switching, Python, Git/Bitbucket.
Links: Resume bullet (HPE).

Project: Automated Visio Diagram Generation API (Charter) (2016)

Problem: Creating network diagrams took many hours per request.
Action: Built a server‑side API (Python + PHP) to parse JSON payloads and render .vsdx drawings automatically.
Result: Cut delivery time from multi‑hour manual work to ~10 minutes per diagram.
Stack: Python, PHP, Visio automation, JSON.
Links: Resume bullet (Charter).

Project: Blue Planet MDSO Service Orchestration (Spectrum) (2019–2020)

Problem: Needed end‑to‑end, zero‑touch lifecycle automation; Cisco ASR9k devices weren’t onboarded.
Action: Trained, developed, and deployed Ciena Blue Planet MDSO; added ASR9k device automation; built Python telemetry to capture CPU/memory for downtime analysis.
Result: Expanded orchestration coverage to ASR9k and improved troubleshooting visibility with scripted metrics collection.
Stack: Blue Planet MDSO (SDN/NFV), Python, Cisco ASR 9000.
Links: Resume bullets (Spectrum Developer II).

Project: Access‑Control Hardening & Outage Prevention (Spectrum) (2017–2020)

Problem: Misused privileges caused trunking outages and inconsistent device changes.
Action: Implemented command restrictions in Cisco ACS/TACACS+; standardized change inputs with web JS forms; enforced versioning with Git/GitLab.
Result: Prevented trunking‑related outages and raised configuration consistency across teams.
Stack: Cisco ACS, AAA/TACACS+, JavaScript, Git/GitLab.
Links: Resume bullet (Spectrum Engineer II).

Project: Data‑Center Turn‑Up via Version Control (Amazon/Apex) (2022–2023)

Problem: Siloed device data and manual provisioning slowed host/TOR bring‑up.
Action: Verified and uploaded network data to AWS repos; configured Linux hosts and Juniper TORs using Python in a DevOps workflow.
Result: Centralized, auditable configs and faster, more consistent deployments.
Stack: Linux, Juniper, Python, AWS version control.
Links: Resume bullet (Apex @ Amazon).

Project: Cisco ACI Lab Build & Readiness (Arm/Hays) (2022)

Problem: Internal ACI lab needed end‑to‑end setup and operational readiness.
Action: Racked/stacked hardware; configured via ACI GUI; earned 40 Cisco CE credits; integrated into a global follow‑the‑sun operations model with proactive monitoring.
Result: Delivered a working ACI lab for global teams with improved alerting/hand‑offs.
Stack: Cisco ACI (underlay/overlay), DC ops.
Links: Resume bullet (Arm).

Project: PatternPals — Juggler Match & Scheduling App (2025)

Problem: Juggling partners lacked a way to match by skills/availability and plan sessions.
Action: Built a cross‑platform app with offline‑first sync, auth, real‑time notifications, and a searchable pattern library; migrated from mock data to a Supabase backend with RLS.
Result: Established a multi‑user cloud backend and completed authentication/notification flows with documented tests and migration notes.
Stack: React Native (Expo), TypeScript, Supabase/Postgres, React Navigation, AsyncStorage.
Links: GitHub — pattern‑pals.

Project: Simple Harmony Maker — Web Accompanist (2024–2025)

Problem: Musicians needed on‑demand harmonic context without a human accompanist.
Action: Shipped an interactive web app with chord/roman‑numeral entry, keyboard‑based chord builder, metronome, loop/tempo/time‑signature controls; launched subscription via Stripe; deployed on Vercel and a custom domain.
Result: Public release with full feature set and subscription flow ($10/month) using Stripe checkout.
Stack: JavaScript/HTML/CSS, Tone.js/Web Audio API, Stripe, Vercel.
Links: App — simpleharmonymaker.com • GitHub — simple-musical-accompaniment.

Project: Browser Metronome (2025)

Problem: Needed a minimal, hackable metronome for browser practice and audio‑UI experiments.
Action: Implemented a lightweight metronome using the Web Audio API; structured for easy extension and experimentation.
Result: Reference app suitable for timing/audio UI work in the browser.
Stack: TypeScript, HTML, CSS, Web Audio API.
Links: GitHub — metronome.