Lucas Caldas
PT
Architecture & Backend · Automation & AI · Cloud & GCP

Building reliable systems.
Solving complex problems.

Lucas Caldas

I build with quality, earn trust through consistency, and solve hard problems with clear and reliable execution.

Backend EngineeringAI Agents & AutomationCloud Architecture / GCPMLOps & Data Pipelines
See Delivery Cases Architecture Approach
Delivery Pipeline
WARMING

Outcome Focus

Reliable backend systems and automated decision flowsArchitected for scale, observability, and fast iteration

Core Expertise

  • ›Backend Engineering
  • ›AI Agents & Automation
  • ›Cloud Architecture / GCP
  • ›MLOps & Data Pipelines

Active Stack

Python · C# · TypeScript

GCP · Vertex AI · ADK

MLOps · Airflow DAGs · Apache

Describe your challenge and discover relevant solutions.

Scroll
Domains

Areas of expertise

01

Backend Engineering

Java, Python and .NET/C# services with resilient APIs, asynchronous workloads, and high-availability architecture.

See example
02

AI Agents & Automation

Multi-agent workflows with ADK, Vertex AI, and MCP integrations to automate business operations end-to-end.

See example
03

Cloud Architecture / GCP

Cloud Run, BigQuery, Pub/Sub, IAM, and serverless-first platform design that scales with predictable cost and reliability.

See example
04

MLOps & Data Pipelines

Orchestrated pipelines with Airflow DAGs, model lifecycle operations, and production-ready ML systems that stay observable.

See example
Selected Projects

Featured work

All projects
AI Agents

AI Agents with ADK

Autonomous agents integrating Jira, Looker, and BigQuery via MCP Toolbox, deployed on Cloud Run with Vertex AI.

PythonADKVertex AI
Case study
Payment Systems

Payment Integration Platform

Full Stripe integration: webhooks, automated checkout, subscription management, and failed payment recovery.

PythonStripeWebhooks
Case study
Payment Systems

GPOS Payment System

Complete .NET/C# payment system with PIX, cards, mTLS, and remote terminal registration, eliminating the need to ship physical devices.

.NETC#PIX
Case study
Labs

Technical demos, live.

Labs is where architecture becomes interactive. Agentic workflows, data pipelines, and system simulations shown as running systems, not slides.

Explore Labs

Multi-Agent Explorer

View lab

Compare traditional and agentic workflows solving the same task with different execution models.

MCP Explorer

View lab

Inspect an agent request as it moves through MCP server integrations and external systems.

Cloud Architecture Explorer

View lab

Interactive architecture canvas for modern event-driven systems on Cloud Run and BigQuery.

About

I work at the intersection of GCP cloud architecture and AI agents with MCP, grounded in shipping backends and integrations under real production pressure. I prefer systems that explain themselves through behavior, not only diagrams.

How I build

Let's talk

Available for senior engineering positions and architecture consulting.

Get in touch