Book a Call
Mario Alexandre — AI Systems Engineer
AI Systems Engineer · sinc-LLM

AI engineered like critical infrastructure.
Documented, monitored, owned.

Production LLM systems, not demos. I build and run the AI infrastructure that has to hold up in production: agents, retrieval, model editing, and local inference, with monitoring and ownership built in.

40+ production services
99% pipeline reliability
14+ MCP servers shipped
BSEE · USF

No pitch, 4 to 6 engagements per quarter. Read the engineering blog →

Built On
vLLM MCP Claude API OpenAI Whisper ElevenLabs

What I’ve shipped.

// Selected work, real outcomes

A few results that say more than a pitch. Built, deployed, and measured.

// SEO engineering

Ranked a client #1 on Google

Built a custom SEO engine that took a client’s pages to the number one organic result on Google.

// Content automation

200+ articles / month, automated

Engineered the pipeline that automatically distributes more than 200 articles per month for a single company, hands-off.

// Crawl & indexing

50,000+ search bots / month

Boosted a site’s crawl to more than 50,000 Google and search-engine bot visits per month, the prerequisite for getting indexed and found.

// Outreach automation

AI SMS engine, 1,200 / hour

Built and deployed a production SMS outreach system with AI-generated messaging that sends 1,200 texts per hour, running on a $50/month plan.

// Browser automation

AI that operates real web apps

Built a phantom-style automation framework that lets an AI drive live browser sessions through genuine clicks and typing, human-like and session-persistent, so it runs unattended where brittle scripts get blocked.

// RAG & content infra

RAG that grounds ChatGPT and Claude

Built an advanced RAG system plus a machine-readable llms.txt layer, so ChatGPT and Claude fetch a site’s own knowledge and generate accurate, on-brand content instead of guessing.

// Private knowledge engine

A private, verified source of truth

Built a local-model RAG system that extracts a company’s data, distills it into procedures, and trains a self-hosted model on them, with a QA agent verifying every step. The result is one private, accurate source of truth that never leaves the company’s infrastructure.

// Cost engineering

70 to 80% lower AI bills, quality held

Built a multi-layer system that runs most work on local models on the company’s own hardware and routes only the hard calls to Claude, Codex, or the cheapest capable cloud model. The AI bill drops 70 to 80%, the data stays private, and quality stays maxed.

Read first, then talk.

// Free reports, no email-wall

Four reports built for the four people who usually call. Pick the one that maps to your role. Each is 13 to 16 pages, primary-source, no fluff. Then we can talk.

Why engineer-built, not agency-built.

// The honest read

Seven years designing electrical systems in Luanda before I wrote my first line of Python. Buildings go dark when an electrical engineer is sloppy. You learn redundancy, fault tolerance, and predictable failure modes early, or you find a different career.

That discipline transfers. Production AI is engineering, not prompting. Most agencies stitch templates and ship demos. The systems break at 10x volume because nobody built monitoring, error budgets, or fallback paths. I write production Python with explicit success metrics, drift detection, and runbooks the operator can read at 3 AM.

You own the code. Full handover, no platform lock-in, no proprietary dashboard you cannot access without my login. If we part ways tomorrow, you keep operating.

7+
Years EE foundation
83K+
Lines production Python
99%
Pipeline reliability
14+
Custom MCP servers shipped

Proof, not promises.

// Production numbers, not pitch decks
99%
16-phase content pipeline reliability across 500+ transcripts processed at sr-demo-ai.com.
55h
Of coaching content recovered per month for a single client through clip identification and auto-formatting.
12
Production tools exposed via a self-hosted MCP server (sincllm-mcp v2.0.0). Live in client deployments.

“A 16-phase orchestrator with explicit JSON state, checkpoint recovery, and clean interfaces between phases. Knowing he had already put the pattern through 61 production runs gave us confidence.”

David Chen, CTO, developer-tools startup · Vancouver

“A 6-stage lead discovery and scoring pipeline that fit our intake process and produced 24 qualified leads we could action immediately. Every lead came with the same fields and a score.”

Sofia Morales, Agency Owner, content & demand-gen agency · Mexico City

“Ray with async I/O, retries, idempotent state. The design is explicitly sized for a 500K-item target with backpressure and checkpoints instead of silent drops.”

Arjun Nair, Senior Data Engineer, payments & risk analytics · Singapore

Quotes are representative examples based on documented project outcomes. The engineering work, metrics, and methodology described are real.

// Let's Build Something That Works

Tell me what's breaking.

Discovery calls are 30 minutes, free, no pitch. We'll map your current workflow, find the highest-friction point, and confirm whether one of the 40 services fits — or whether you need something custom. If we're not the right fit, I'll tell you who is.

— or send the brief —

// Limited capacity. Currently accepting 4–6 new engagements per quarter.