Dani Roxberry
AI Technologist | Applied AI Leader | AI Platform & Agent Systems | Technical Founder
Greater Boston, MA
512-850-5067 | droxey@gmail.com
Summary
Applied AI technologist, technical founder, 2x CTO, professor, and platform builder with 22+ years shipping production software and 9 years helping engineers turn new technology into working systems. I build the bridge between AI strategy and the real work: architecture, code, operations, governance, adoption, and measurable business value.
Built 4 startups, hold 1 U.S. patent, invented live-streaming video conversation technology, and taught in Make School's Y Combinator-backed accelerator environment. I write production-quality code, design AI systems, lead technical programs, and explain complex systems clearly to founders, executives, engineers, and non-technical stakeholders.
Hands-on with Claude Code, Anthropic API, MCP, RAG, OpenClaw, LiteLLM, multi-agent workflows, AI code review, LLM evaluation, safe self-hosted AI agents, and production AI deployment. Strong fit for AI leadership roles that need technical judgment, responsible AI practices, developer adoption, and the ability to move from pilot to production.
Core Skills
- AI Platform & Operations: MLOps, LLMOps, model lifecycle, AI observability, deployment verification, smoke tests, guardrails, cost controls, platform reliability, production AI deployment, incident-aware operations
- AI Strategy & Leadership: AI strategy, applied AI roadmaps, AI transformation, AI product strategy, AI adoption, executive communication, stakeholder alignment, technical decision-making, business outcomes, responsible AI
- AI Systems & Architecture: LLMs, generative AI, RAG, agentic workflows, multi-agent systems, AI-native applications, AI infrastructure, model routing, tool use, memory systems, evaluation loops, prompt systems
- Developer Enablement & Adoption: workshops, technical office hours, code review, tutorials, docs, runbooks, hackathons, build-a-thons, demo apps, working prototypes, public speaking, mentoring, technical curriculum
- Engineering & Cloud Architecture: Go, Python, TypeScript, JavaScript, C#, SQL, React, Node.js, Django, REST APIs, distributed systems, cloud-native architecture, microservices, event-driven systems, concurrent systems
- Governance & Security: OWASP, secret handling, egress controls, network isolation, API key handling, secure deployment patterns, auditability, privacy-aware design, safe defaults, responsible AI practices
- Infrastructure & Data: Docker, Kubernetes, Ansible, Linux, Git, CI/CD, Infrastructure as Code, PostgreSQL, MongoDB, Redis, DuckDB, MySQL, SQL Server, Qdrant, Caddy, Prometheus, Grafana
- Startup & Product Leadership: technical founder, 0-to-1 product development, product-market fit, platform adoption, founder credibility, team scaling, open-source communities, investor-ready demos, customer feedback loops
Selected AI Leadership Results
- Built a hardened AI deployment system using Anthropic API support, LiteLLM model routing, Smokescreen egress controls, Docker networking, smoke tests, CI, and OWASP-aligned security practices.
- Designed and taught a 2026 AI engineering course where students built with Claude Code, Codex, MCP, RAG, agent workflows, verification loops, durable repo context, tool use, and production code review.
- Built AI code review pipelines processing 200+ submissions weekly, reducing evaluation turnaround by 80% and cutting course development time by 50%.
- Activated 60+ engineers across 5 courses and twice-yearly hackathons guiding builders from first AI workflow to prototypes, deployment workflows, and public demos.
- Delivered 500+ technical talks, live coding sessions, architecture walkthroughs, and debugging sessions across applied AI, backend engineering, deployment, security, and software engineering.
- Ran 10+ hackathons and build-a-thons where cross-functional teams moved from idea to MVP, deployment, and live demo within a single event cycle.
- Founded 4 startups and invented patented live-streaming video conversation technology, taking a hard technical product from prototype through production release and platform scaling.
- Designed a 16-course applied computer science program covering AI engineering, data structures, web security, DevOps, software engineering, and production software architecture.
- Published open technical systems and curriculum across personal, company, and university repositories; course repositories earned 60+ GitHub stars and 100+ forks.
Selected Applied AI Projects
- Clincher: Automated secure deployment harness for self-hosted AI agents using Ansible, Docker Compose, Smokescreen egress allowlisting, LiteLLM routing, Redis semantic cache, Caddy, Cloudflare, sandbox hardening, verification commands, and OWASP-aligned deployment practices.
- GrainDL: Production-grade Go CLI for exporting Grain meetings, transcripts, highlights, AI notes, metadata, audio, and video into portable archives with headless Chromium automation, Rod, ffmpeg, markdown output, Docker support, retries, rate limiting, and structured manifests.
- Skillsport: Curated AI agent skills and a safe-by-default CLI for scanning, auditing, planning, fixing, migrating, validating, and semantically searching reusable skill files across repositories.
Professional Experience
Muse & Machine - Founder / Principal Applied AI Engineer
Remote | Jan 2023 - Present
- Built production infrastructure for self-hosted AI systems on Ubuntu 24.04 VPS servers; automated model routing, outbound network controls, vector storage, reverse proxying, monitoring, backups, health checks, and verification commands.
- Hardened AI-agent deployments with OWASP-aligned controls: egress allowlisting, blocked-egress tests, secret isolation, Docker network boundaries, Redis credential handling, HSTS/CSP headers, and repeatable security checks.
- Integrated Anthropic API and Claude model routing through LiteLLM and OpenClaw, including API key handling, model registration, semantic-cache fixes, smoke tests, and failure-mode documentation so builders can verify real access.
- Used Claude Code, Claude PR Assistant, and Claude code-review workflows to ship fixes across security hardening, Docker networking, egress controls, secret handling, CI, Molecule tests, ShellCheck, pre-commit, and deployment reliability.
Dominican University of California - Professor of Applied Computer Science
San Rafael, CA | Aug 2021 - Present
- Led a hands-on AI engineering curriculum where students used Claude Code, Codex, MCP, RAG, agent workflows, repo instructions, tool calls, verification loops, and code review to ship agent-assisted features.
- Built AI code review pipelines processing 200+ submissions weekly and onboarded students and faculty into the workflow; reduced feedback turnaround by 80% and course development time by 50%.
- Activated 60+ engineers across 5 courses and twice-yearly hackathons, moving builders from technical instruction to working prototypes, deployment workflows, and public demos.
- Led hundreds of engineering teams through developer onboarding, sprint planning, architecture reviews, code review, deployment, and demo readiness; held teams to production-quality shipping standards.
- Designed and delivered 4+ applied computer science courses per semester in AI engineering, Go, Docker/DevOps, web security, data structures, and JavaScript for cohorts of 10 to 60 junior engineers.
- Built workshops, tutorials, labs, and technical content for an applied CS program; published materials through GitHub with CI and automated evaluation.
- Coached engineers through debugging, architecture, execution, and interview prep; alumni shipped code at Netflix, Meta, Microsoft, Oracle, and Tesla, and several founded $1M+ revenue startups.
- Turned repeated student and team friction into better onboarding docs, starter repos, checklists, rubrics, and feedback loops.
Make School - Instructor, Full-Stack Web
San Francisco, CA | Aug 2018 - Jul 2021
- Taught in Make School's Y Combinator-backed accelerator environment, preparing early-career developers and technical founders to ship full-stack products under production constraints.
- Ran 6+ backend and full-stack courses per semester covering Python, Node.js, APIs, and databases.
- Developed 12+ developer enablement modules focused on hands-on labs, test-driven development, peer review, code quality, and production engineering.
- Coached 250+ junior engineers through debugging, portfolio projects, technical interviews, and deployable software.
- Automated slide and lesson plan generation, freeing 50% of prep time for direct mentoring, technical coaching, and workshop delivery.
- Grew open-source developer mindshare through course repositories that earned 60+ GitHub stars and 100+ forks.
UC Berkeley Extension - Instructor, Full-Stack Web / Data Science
San Francisco, CA | Jul 2017 - Dec 2018
- Taught full-stack web development and data science/visualization through repo-backed technical docs, live demos, project specs, debugging sessions, and deployment walkthroughs.
- Ran deployment workshops that moved 200+ aspiring engineers from classroom exercises to shipped capstone projects on production infrastructure, including Heroku deployment, CLI workflows, API key handling,
.envsecret management, GitHub-based releases, and real user feedback. - Developed open-source course documentation for Python 3, Django, virtualenv, Jupyter Notebook, conda environments, VS Code, Plotly, R tooling, Git, GitHub, SSH configuration, and networking basics.
- Delivered hands-on technical sessions that moved developers from curiosity to active building through Django scaffolding, migrations, browser testing, deployment, product iteration, and demo readiness.
FrameBuzz - Founder / Inventor / Chief Technology Officer
Austin, TX | Oct 2012 - Dec 2017
- Founded and led a 0-to-1 startup product from prototype through production release and platform scaling.
- Invented and patented live-streaming video technology, the core IP behind FrameBuzz.
- Built the video annotation engine from the ground up and led 4 engineers across architecture, implementation, release, scaling, and daily operations.
- Pioneered timestamp-anchored viewer interaction for live and recorded video years before comparable engagement features appeared in mainstream video platforms.
- Built founder credibility by owning product, engineering, investor conversations, customer feedback, technical storytelling, and public demos from idea through production.
Earlier Experience
bitoriented inc. - Founder / Principal Software Engineer
Remote | Nov 2014 - Aug 2018
Delivered custom software across Django, PHP, .NET, iOS, e-commerce, telecom, healthcare, education, and media. Turned vague business requirements into scoped, shippable production systems and helped non-technical founders understand what was being built, why it mattered, and how to keep it running.
ProAce Technology Services - Senior Systems Engineer
Austin, TX | Apr 2011 - Mar 2013
Shipped 8+ client projects across iOS, Django, PHP, and .NET for e-commerce, telecom, and healthcare clients. Owned scoping, implementation, testing, launch, and production deployment across concurrent client engagements.
HZDG - Senior Interactive Developer
Remote | Oct 2010 - Jan 2011
Joined after HZDG acquired Titrate Solutions; mentored developers, ran code reviews, and helped accelerate delivery.
Titrate Solutions - Chief Technology Officer
Austin, TX | Sep 2009 - Dec 2010
Led a 3-person engineering team that launched 12+ web, interactive, and mobile products in one year for NFL, IMAX, Brooks Brothers, and other major brand clients.
NiCE - Junior Software Engineer
Midvale, UT | Apr 2008 - Sep 2009
Collaborated on the next-generation inContact product, including client-facing call-center features, administrative tools, and application alerting for the Network Operations Center. Led a student intern team working with analysts, developers, and interface designers in a fast-paced Agile environment.
Education
Neumont College of Computer Science - BSCS, Computer Science
Salt Lake City, UT | 2006 - 2008
- Completed bachelor's degree in 2 years.
- Charter Member, Society of Women Engineers: 2007
Patents
US20140199046A1 - Conversations on Time-Shifted Content
Invented live-streaming video technology for timestamp-anchored conversations and viewer interaction on live and recorded media.
Open Source
- GitHub: Dani Roxberry: Public engineering work across AI systems, deployment, prompt infrastructure, technical curriculum, and developer onboarding.
- GitHub: Muse & Machine: Applied AI product experiments, multi-provider LLM infrastructure, and secure self-hosted agent operations.
- GitHub: Applied Computer Science Curriculum @ Dominican University: Open-source course materials, assignments, CI, automated evaluation, and applied software engineering education.