Alan M.Private profile: Upwork shows its earnings only to signed-in members, so we check it every few weeks, not daily. MRR here is a rough estimate (Upwork's six-month earnings ÷ months active). Last read Nov 5, 2026.

Most AI projects do not fail on the model. They fail on the server it runs on, the API keys nobody rotated, and the alert that never fired at 3am. That part is my job. I have spent 15+ years running production Linux systems, and 1,800 hours of it here on Upwork at 100% job success. I design, harden and automate Linux infrastructure on AWS, Google Cloud and DigitalOcean. These days that includes the infrastructure AI runs on: deploying agent workflows built with n8n, LangChain and the OpenAI and Claude APIs, then locking down the keys, the logging and the recovery path so the thing still works next month. I also build and run a public MCP server and REST API of my own, so I know what it takes to keep agent tooling live for other people's systems. WHY CLIENTS HIRE ME 1. Proven enterprise trust (Top Rated): large, successful contracts; consistent 5-star outcomes 2. Operational excellence: 99.9%+ uptime architectures, zero-/near-zero-downtime migrations, robust rollback plans 3. Real savings: right-sizing, consolidation and automation that measurably cut client infrastructure bills 4. Security first: layered hardening (WAF, iptables, DNSSEC), least-privilege access, audit-ready logging WHAT I DELIVER - AI Systems in Production: deploying and hardening agent workflows built with n8n, LangChain and the OpenAI and Claude APIs; API key handling, least-privilege access, logging and a tested recovery path. Python and REST API integration work - DevOps & Automation: CI/CD, immutable builds, environment parity; Ansible & Bash at scale - Cloud & Containers: design/ops on AWS | GCP | DigitalOcean; Docker & Kubernetes (incl. k3s) for resilient apps - Web & Networking: Nginx/Apache tuning, HA, SSL/TLS, Bind9/DNS, HAProxy load balancing - Email & Deliverability: Postfix, Zimbra, Mautic, Postal; SPF/DKIM/DMARC alignment and sending infrastructure - Monitoring & Reliability: CheckMK and Zabbix visibility, alerting that reduces MTTR Message me with your current stack and goals, I’ll propose a lean plan that reduces risk, controls cost, and gets you to green faster.