AI and cloud engineers who run what they build
We are AI and cloud engineers. We build AI into real systems — retrieval over your own data, Claude Skills and plugins, agents, whole AI products — and prove the patterns in our own live products first, on infrastructure we design ourselves. The tools on this site are the parts of that work that turned out to be worth automating, built because we needed them, not as a product strategy.
AI implementation
RAG that answers from your own documents with citations, Claude Skills and MCP integrations that give AI your team's tools, and agents with a person in the loop when confidence is low. The patterns run in our own live products first: cited answers and human review in AetherERP, a tool-using assistant in Stock Tracker, a guarded AI rewrite in Tally.
Cloud architecture
Terraform, managed identities, private networking and CI that plans on a pull request and applies on merge, on AWS, Azure and GCP. The infrastructure behind this site and our own products is built the same way.
Production discipline
AetherERP has token budgets, a log of every model call and a fallback when a model is unavailable. Tally's AI rewrite falls back to the template if it changes a fact. Our deploys check the new version is actually serving before they report success. And we use our own tools: the IaC reviewer runs against the Terraform behind this site.
Why trust us
Industry Certifications
AWS Solutions Architect, Azure Solutions Architect Expert and Google Professional Cloud Architect — held, not aspired to.
Notable Projects
Cloud and infrastructure work for teams at Prime Therapeutics, UHG Optum, Change Healthcare, Carnival, Raytheon and FCMAT, delivered directly and through contract and subcontract partners. It set the production standard our AI work is held to.
AI that admits its limits
AetherERP's policy assistant discards any citation it did not actually retrieve. The App Generator reports a build it could not fix rather than hand over code that does not compile, and every run, failures included, is public on the benchmark page. We would rather lose a demo than a deployment.
How we engage
A free first review, then fixed-scope delivery (a RAG system, a set of Skills and integrations, an agent, a migration) or engineers alongside your team month to month. Pick whichever fits what you are actually trying to finish.
Who you would be working with
Robert Alfaro
Founder — AI implementation, cloud architecture and platform engineering
Two decades of building and running cloud infrastructure for healthcare, defense, travel and public-sector organisations — the engagements named above. Now the work is putting AI into production systems. That means the retrieval, agents and AI features inside AetherERP, Stock Tracker and Tally. It also means building Claude Skills and MCP integrations for teams that want AI working inside their own tools. The tools on this site started as the parts of all that work worth doing once instead of every time.
Ask where AI fits — it costs nothing
Tell us the process you want AI to take on, or send the architecture or Terraform you are unsure about. Within a few days you get a short written plan: what we would build or change, what it would take, and what to avoid. No call required. Or try the App Generator and watch it prove its code on real CI.
We reply within one business day.