Personal technical portfolio

LABP-Stack

A hands-on lab connecting a foundation in voice and real-time communications with growing work in automation and applied AI.

Luis Alfredo Báez Páez
Technical Infrastructure & Communications Leader

Evidence-first status: this portfolio contains a static site, small tested simulators, a keyword-based API prototype, and clearly labeled design studies. It is not a production platform or commercial service offering.

Current work

Prototype · deployed demo

Intent API

Node.js/Express endpoint using deterministic keyword rules for PRICING, GREETING, and UNKNOWN. It includes request IDs and structured console logs; it does not use an AI model.

Inspect backend code

Implemented · tested simulator

Voice AI Gateway

A reproducible Node.js decision layer that classifies transcript events and returns explicit queue, workflow, or clarification actions. Live ARI and audio are not included.

Open project case study

Implemented · credential-free simulation

Change Request Automation

An importable n8n workflow validates fictional change requests, routes them to human review, prepares committee summaries, and records structured outcomes without external credentials.

Open automation case study

Planned

RAG experiment

Concept scope only. No ingestion, embeddings, vector store, retrieval, generation, or evaluation code is currently committed.

Read planned scope

Current architecture

Browser text input
      -> POST /analyze
      -> Express API
      -> keyword rules
      -> JSON intent and response

AWS, live Asterisk/SBC integration, speech-to-text, and RAG remain roadmap or design-study items. The published n8n artifact is an isolated simulation, not a production integration.

API demo

Checking the prototype backend…

Next milestones

Keep the backend API contract and automated tests aligned with the deployed prototype.
Document deployment limits and operational observations from the hosted demo.
Add an isolated ARI adapter to the tested Voice AI Gateway decision layer.
Capture a local n8n execution of the credential-free change-request simulation.
Implement and evaluate a small RAG experiment before describing it as a working component.

Current work

Intent API

Voice AI Gateway

Change Request Automation

RAG experiment

Current architecture

API demo

Next milestones

Source and design notes