Skip to content
View SAMIRRICARDO's full-sized avatar

Block or report SAMIRRICARDO

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
samirricardo/README.md

Samir Ricardo

AI Systems Engineer · AI Systems Architect · Founder of VRASHOWS

LinkedIn Email Book


I build production AI systems: autonomous agents, multi-layer RAG pipelines, evidence-based observability, and multi-tenant cognitive platforms. 15 years in software engineering, focused on AI infrastructure since 2022.


VRAXIA WORK — Autonomous Job Application Engine

Fully autonomous job application system built on TypeScript + Playwright. Applies to positions, fills forms using a 5-layer answer resolution pipeline, captures multi-source evidence, and audits every application with an independent verification engine.

Architecture highlights:

  • ApplicationStateMachine — enforces lifecycle transitions at runtime; illegal state transitions throw immediately
  • ApplicationTruthEngine — evaluates evidence from network captures, screenshots, DOM snapshots, and platform confirmations independently from workflow state
  • EvidenceCollector — multi-source evidence capture per application for post-hoc auditability
  • ApplicationTracer — structured observability with per-step timing, retry counts, and error classification
Metric Result
Applications submitted 82+ across 3 platforms
Evidence coverage 100% of confirmed applications produce a structured TruthRecord
Error categories 15 categories with RCA and configurable retry policy

github.com/SAMIRRICARDO/vraxia-work-system


VRAXIA Sense — Proactive AI Perception Layer

"Traditional AI responds when you ask. VRAXIA Sense perceives before you notice."

Three-tier event processing pipeline that monitors external signals and acts autonomously before a human opens the inbox.

Metric Target Production
Level 0 noise elimination >70% 76%
Avg. cost per raw event <$0.001 $0.000034
Event-to-notification latency <2 min ~8 seconds
False positive rate <10% ~6%

github.com/SAMIRRICARDO/vraxia-sense


Human RAG

Traditional RAG indexes what was written. Human RAG indexes how a person thinks and decides — reconstructing reasoning patterns, not just retrieving text.

Implemented in VRAXIA OS: enterprise platform with 8 departmental AI agents, 1,100+ indexed skills, and multi-tenant BYOK architecture.


Stack

TypeScript Node.js Claude PostgreSQL Redis Docker Playwright

ESM · Express + SSE · Local RAG (JSONL + TF-IDF) · BYOK multi-tenant
Token cost tracking · Dynamic model routing · Per-domain batch governance


Publication

O Maior Ativo da Sua Empresa — Amazon KDP · June 2026
The first Brazilian book on Human RAG applied to organizational knowledge preservation.


VRASHOWS · contato@vrashows.com.br

Popular repositories Loading

  1. samirricardo samirricardo Public

    Portfólio Samir Ricardo

  2. vraxia-work-system vraxia-work-system Public

    Autonomous job application system: FSM lifecycle enforcement, 5-layer questionnaire resolution, evidence-based verification, real-time dashboard

    TypeScript

  3. vraxia-system-os vraxia-system-os Public

    AI operations dashboard for VRAXIA OS: Next.js 15, real-time agent monitoring, multi-department KPIs

    TypeScript

  4. vraxia-showcase vraxia-showcase Public archive

    AI Operating System for Enterprise — Portfolio Showcase

  5. vraxia-sense vraxia-sense Public

    Proactive AI perception layer: 3-tier event classification at $0.000034/event with 76% zero-cost noise elimination

    TypeScript