01 / Developer toolsOpen source
SAM Doctor
Deployment failures leave clues in different places. SAM Doctor brings the logs, templates, and CloudFormation evidence into a diagnosis you can follow.
AWS · Python · CI/CDView source
IT support & infrastructure · Fairfax, Virginia
Jacob
Goldstein.
IT Support Specialist at The Potomac School.
I build practical tools and write down what I learn along the way.
A few things I’m working on
On the workbench.
02 / Historical recordsOpen source
AKT Reader
A local reader for handwritten Polish and Cyrillic registers, keeping the original evidence, uncertain readings, and review state close together.
Python · Handwriting · ResearchResearch tools
Ticket review workflows and Mac printer verification.
The useful paper trail
Notes from the work.
Problems, fixes, and the part worth remembering.
FN-001The migration that couldn't dropPitt Visual Media Workshop · summer 2025
Situation
A research team's web presence — WordPress Multisite, SFTP-served projects, custom DNS, SSL — had to move, and the team had to keep working through the move.
Approach
Validated hardware, SSL and DNS paths, and deployment targets before each move rather than after; migrated in stages; chased down the permissions and deployment issues each stage surfaced.
Outcome
No unplanned downtime across the migrations, and a written handoff so the next person didn't need me in the room.
Lesson
Unplanned downtime gets scheduled earlier than anyone thinks — by whoever skipped validation.
FN-002The 2.5 GB model that loaded every cycleMailBridge · personal project
Symptom
My local email-triage pipeline pegged the CPU on every polling cycle, far past what fresh mail volume could explain.
Diagnosis
Two compounding causes: summaries were recomputed for messages that hadn't changed, and the 2.5 GB local model was reloaded from disk every cycle instead of living across them.
Fix
Cache summaries by message_id; fix the model lifecycle so it loads once. Each fix shipped as a written spec with acceptance tests, then verified by a one-command harness before it counted as done.
Lesson
Measure before optimizing — the expensive thing is rarely the thing you assumed.
FN-003Deploy failures that hid their evidencesam-doctor · personal project
Symptom
AWS SAM deployments through GitHub Actions fail with errors that name the symptom but not the cause — and the cause is spread across templates, build logs, and CloudFormation events nobody reads under pressure.
Response
Built a local diagnostic tool that reads the actual logs and templates and refuses to guess: every diagnosis it offers cites the evidence it was drawn from.
Lesson
An answer without evidence is a hypothesis. Tickets close faster when the diagnosis shows its work.
A little context
Curiosity, on
and off the clock.
I work in IT support at The Potomac School. Before that, I studied computer science at Pitt and worked on research infrastructure, web migrations, and AWS deployments.
The part I keep coming back to is understanding how the pieces fit: the person, the device, the network, and the software. My side projects follow the same curiosity into home labs, local tools, and historical records.
Away from the screen, that usually means a hike or another PC project.
Find me on GitHub
Experience & background
Experience
IT Support Specialist
The Potomac School
Support across endpoints, identity, and access in a school environment, backed by methodical troubleshooting and clear documentation.
Systems & Infrastructure Technician
University of Pittsburgh — Visual Media Workshop
- Provided technical support and infrastructure assistance for a specialized research team — Linux systems, Apache configuration, DNS changes, SSL transfer, SFTP migration, and a WordPress Multisite migration.
- Ran migrations without unplanned downtime by validating hardware, SSL/DNS, and deployment paths ahead of each move, and troubleshooting permissions and deployment issues.
- Monitored NVIDIA GPU workloads for AI training and performed daily throughput benchmarking.
- Automated retrieval and restructuring of hierarchical data via Python to preserve researcher data from Itinera, the team's digital-humanities project, so its visualizations would survive off the live server. Documented the work for handoff (prototype explorer).
System Developer
Pricing Empowered
- Deployed and supported 4 AWS Lambda services across 3 environments via AWS SAM, cutting deployment time from hours to minutes.
- Documented deployment steps and collaborated with front-end teams to troubleshoot API and latency issues.
- Integrated RESTful API endpoints and supported staging/production release workflows with Python and SQL.
- Built GitHub Actions / AWS SAM CI/CD pipelines, reducing recurring deployment errors.
Capabilities
Support & Troubleshooting
IT troubleshooting, user support, hardware troubleshooting, ticket and technical documentation, escalation, Windows fundamentals, customer support.
Systems & Networking
Linux, DNS, SSL/TLS, networking fundamentals, Apache, WordPress, SFTP, permissions, NVIDIA GPU workstations.
Cloud & Automation
AWS (Lambda, S3, SAM, CloudFormation), Docker, Kubernetes fundamentals, Bash, Python, SQL, Git, GitHub Actions, CI/CD, REST APIs.
Education & certification
B.S. Computer Science Cum Laude
University of Pittsburgh — Class of 2025
Secondary field: Applied Mathematics & Statistical Computing.
Coursework included robotics and ROS — academic experience.
M.S. Applied Information Technology Accepted
George Mason University — Cyber Security concentration
Part-time; start term not yet set.
More projects & experiments
Local, evidence-first diagnostics for AWS SAM, CDK, CloudFormation, and GitHub Actions deployment failures — the tool I wished for while running SAM deployments at my internship. Works from the real logs and templates, and every diagnosis cites the evidence it was drawn from.
Provisioned local clusters with Kind and Bash; scripted lifecycle management to monitor pod networking and service discovery; ran Docker workloads under deliberate CPU/RAM constraints — and documented the setup steps and failure points for repeatable troubleshooting.
Open-source fall-detection prototype running in real time from a single RGB camera. Applied state-machine logic to filter inference noise and reduce false-positive triggers; documented setup and testing for reproducible deployment on the device.
Open benchmark corpus for handwritten-text recognition on 19th-century Jewish vital records from Pułtusk and Serock (Polish and Russian-Cyrillic). Grew out of tracing my own family. Evaluation reports per-page error distributions and name-level accuracy — because a mangled surname becomes someone's wrong ancestor. The evaluator is runnable; the corpus is in preparation and no corpus data is published yet.
MailBridge
Local-first email triage pipeline: read-only Gmail ingestion, offline LLM summarization with per-message caching, a desktop dashboard, and a one-command verify harness. Built spec-driven — each change ships with its acceptance tests, and a skipped test is never reported as a pass.
Caption Semantics & Diffusion Models
Fine-tuned Stable Diffusion v1.5 on a 6,737-image dataset and presented GPU workload performance findings to the Pitt research cohort.
Robotics & ROS
Studied robotics with ROS (Robot Operating System) as part of my degree — hands-on exposure to full physical systems, where debugging means isolating layers across sensors, software, and hardware. The same discipline I use in IT troubleshooting.
PC Building & Home Lab
I build and upgrade my own machines — component selection, assembly, troubleshooting, OS setup — and run a home lab and network that serve as the testing ground for my Linux and Kubernetes work. Personal projects, not professional trade work; the comfort with tools and the careful hands are real either way.
Keep in touch
Always glad
to talk shop.
Working on a similar problem, or have a thought about one of these projects? I’d be happy to hear it.
jacobgoldstein.cs@gmail.comAbout this site
This site
host aws-lightsail
os ubuntu-24.04
web nginx
tls letsencrypt·auto-renew
brute-force fail2ban
uptime
deployed
monitor uptimerobot — live status ↗
backups timestamped·pre-deploy
Run like production
This page is served by an nginx instance I administer on AWS Lightsail. TLS is Let's Encrypt with auto-renewal across apex and www. External monitoring by UptimeRobot — the uptime history is public.
Deployed, not uploaded
Content ships over SSH through a scripted deploy that takes a timestamped backup before every
replacement, sets ownership and permissions explicitly, and has a written rollback plan.
Config changes never skip nginx -t.
Hardened where it counts
fail2ban jails on SSH and HTTP auth. The private area rides Basic Auth over HTTPS only, with noindex and no-store headers and the credential hash kept outside the web root.
Documented for the next person
The whole setup — layout, deploy procedure, auth notes, rollback — lives in a deployment runbook written so someone who isn't me could operate it.
Analytics here are first-party and cookieless — nothing leaves this server. This is a small production system.
Server watch
Loading current summary
- Blocked probes
- — today
- Blocked probes
- — last 7 days
- Top probe source
- — last 7 days
- Server uptime
- — days
Aggregate nginx security-log summary. Probe counts are requests intercepted by explicit hardening rules or suspicious-query guards — not ordinary visitors or IP addresses.