Created a reliability-focused MySQL toolkit covering replication checks, backup and restore drills, and operational validation with clear guardrails. This project demonstrates my production mindset around database reliability, recovery confidence, and safer operations.
1
12
Built a real-time AI interaction architecture using streaming patterns, session-aware communication, lightweight retrieval, and observability-first design. This work highlights my approach to production-minded AI systems where latency, traceability, and reliability matter.
1
14
Designed a practical LLMOps / RAG foundation for grounded AI workflows, with structured ingestion, retrieval, testing, and cloud-ready extensibility. This project reflects how I approach applied AI systems with engineering discipline, not just prototype-level experimentation.
1
17
Built a production-oriented GPU platform on AWS EKS with a strong focus on security, observability, progressive delivery, and cost-aware scaling. This case study shows how I design cloud foundations for AI workloads using Kubernetes, GitOps, policy guardrails, and platform automation.
1
21
Project title:
Chacaltaya
Tagline / One-line description:
A quiet station on the internet. One question, one reading.
Live URL:
https://chacaltaya.falvarezpinto.com
Walkthrough video URL:
https://youtu.be/N6-zkrqydXs
GitHub / Repository URL:
https://github.com/fdaniel-alvarez-dev/chacaltaya
--- SOCIAL PROOF --------------------------------------------------
X / Twitter post URL:
https://x.com/FreddyAlva21905
LinkedIn post URL:
https://www.linkedin.com/feed/update/urn:li:share:7452837542636244992/
--- DESCRIPTION (long) --------------------------------------------
Chacaltaya is a quiet station built on Zo Computer. Visitors ask one question and receive a single, signed reading: one interpretation, one risk, one next honest move.
They can file the reading for the record by entering an email. The site enqueues an outbox entry and a Zo automation fulfills the request using the connected Gmail integration (sends a plain-text copy to the visitor) and the Google Drive integration (archives the same reading as a .txt file in a station folder). Every step writes a restrained entry to the public station log, with no leak of personal data.
Completing a four-stage ascent unlocks a second pass of the Hidden Frequency — a genuine additional transmission, not a badge or a score. Open hours are driven by a Google Calendar integration seam. Payment is handled via a Stripe test checkout that redirects to a station-voice thanks page.
Built with Bun + Hono + React on Zo hosting.
--- INTEGRATIONS / FEATURES CHECKLIST -----------------------------
If the form has a checklist, check all that apply:
[x] Uses Zo hosting
[x] Uses Zo automations
[x] Custom domain via Cloudflare (chacaltaya.falvarezpinto.com (http://chacaltaya.falvarezpinto.com))
[x] Stripe (test checkout via Payment Link, dedicated thanks page)
[x] Gmail integration (delivers reading copy)
[x] Google Drive integration (archives reading as .txt)
[x] Google Calendar integration (seam for live hours layer)
[x] Gamified element (Ascent -> Hidden Frequency unlock, persists via localStorage + public log entry)
[x] Public artifact (reading visible in the station log)
--- TEAM ----------------------------------------------------------
Solo builder:
Freddy Daniel Alvarez Pinto
Location:
Bolivia