Backend Systems Engineer | Infrastructure + AI/ML
📍 Baltimore, MD • 📫 ZackAdamsENG@protonmail.com
I build backend systems, infrastructure tooling, automation workflows, and practical AI applications.
My work sits at the intersection of Python, PowerShell, data engineering, enterprise infrastructure, security automation, and local AI. I am especially interested in systems that ingest messy operational data, detect meaningful changes, expose useful interfaces, and help people make better decisions.
I like software that is:
- Explicit about its assumptions
- Predictable and inspectable
- Honest about failure
- Useful from the command line
- Easy to automate and operate
- Powerful without becoming unnecessarily complicated
Professional systems hacker, more or less.
I currently work on enterprise systems and automation, including:
- PowerShell automation for enterprise workflows
- Microsoft Graph and Entra integrations
- Windows monitoring and change detection
- Vulnerability intelligence ingestion and analysis
- WebAuthn and FIDO2 authentication experiments
- Local AI prototypes for security and infrastructure operations
- PowerShell Universal applications and scheduled automation
- Backend services and REST APIs
- Python and PowerShell automation
- Data ingestion and normalization pipelines
- Monitoring and change detection systems
- Security and vulnerability intelligence tooling
- Semantic search and retrieval systems
- Local LLM assistants and agent workflows
- CLI tools for developers and analysts
- Browser native and peer to peer experiments
- CI friendly validation and reporting tools
A local first vulnerability intelligence platform for ingesting, normalizing, analyzing, and semantically searching the CISA Known Exploited Vulnerabilities catalog.
Includes:
- Fault tolerant PowerShell ingestion
- Raw source snapshots and normalized datasets
- Baseline comparison and change detection
- Structured alerts and reports
- FAISS semantic search
- Optional local querying through Ollama
A PowerShell monitoring system that tracks Microsoft Windows Release Health issues across client and server products.
The system detects changes between recurring runs, maintains a persistent baseline, and produces structured outputs for operational reporting and alerting.
A customized Bitwarden client proof of concept focused on enterprise passkey recognition.
Work includes:
- WebAuthn and FIDO2 behavior
- Custom AAGUID testing
- Browser extension packaging
- Microsoft Entra validation
- Dependency automation
- Maintaining a clean patch set against upstream
A retro desktop gremlin assistant inspired by the downloadable desktop software of the late 1990s and early 2000s.
Kloppy is designed as a local first desktop application with chat history, editable memory, personality settings, and cross platform distribution.
A Python command line tool for diagnosing DNS, TLS, HTTP, and API endpoint problems.
Published on PyPI and designed to turn common connectivity failures into clear, actionable diagnostic output.
A tiny encrypted peer to peer chat application contained in a single HTML file.
AIMless communicates directly between browsers without a traditional backend server.
A browser native distributed computing experiment using WebRTC to coordinate peer to peer workloads.
A local first browser payment SDK for Solana based payments, including wallet connections, balances, token transfers, payment links, and embeddable checkout flows.
A zero configuration Python CLI for auditing CSV datasets before downstream use.
It produces human readable findings, machine readable output, and CI friendly exit codes based on whether a dataset appears safe, questionable, or unsafe.
Languages
Python • PowerShell • SQL • JavaScript • HTML • CSS
Backend and Data
REST APIs • Data ingestion • Normalization • Validation • Structured outputs • Scheduled workflows • Monitoring • Alerting • CLI development
Machine Learning and Local AI
Sentence Transformers • FAISS • Embeddings • Vector search • Retrieval workflows • Ollama • Local inference • Prompt pipelines
Infrastructure
Linux • Windows • Docker • Git • GitHub Actions • CI/CD • PowerShell Universal • Azure
Enterprise and Security
Microsoft Graph • Microsoft Entra • Active Directory • CISA KEV • Vulnerability intelligence • WebAuthn • FIDO2 • Bitwarden
The best system is not always the largest or most abstract.
I prefer small, composable tools with clear inputs, structured outputs, visible state, and understandable failure modes. A useful system should be easy to inspect, easy to automate, and boring enough to trust.
That still leaves plenty of room for weird auth experiments, autonomous agents, serverless peer to peer chat, and browser payment rails.
- Backend and infrastructure engineering
- Enterprise automation
- Security data engineering
- Vulnerability intelligence
- Local AI agents
- Semantic search over operational data
- Authentication and identity systems
- CLI first developer tooling
- Failure aware system design
- Small software with unreasonable capabilities
I am open to engineering roles, interesting infrastructure problems, automation work, AI projects, and collaborations.