About Me

I am Carlos Mora, a Platform Architect with 25+ years of engineering experience spanning physical on-premise data centers, hybrid enterprise backbones, and multi-cloud ecosystems (AWS and GCP).

Over two and a half decades, I have seen the industry cycle through paradigms, tooling trends, and marketing buzzwords. But under production load, the fundamental laws of engineering never change: if your system’s stability, security, or compliance requires engineers to navigate traps with constant human vigilance, your architecture has failed.

My work focuses on eliminating those traps. I design platform architectures where safe, auditable, and cost-controlled operations are the default path. Whether governing distributed data platforms across multi-account Snowflake and Databricks, automating Zero Trust interconnects, or enforcing execution boundaries for agentic AI workloads, I build platforms that protect teams from catastrophic errors while unlocking developer velocity.

Architectural Convictions

  • Deterministic guardrails over human vigilance: A Slack message, a post-deploy checklist, or a “be careful” warning is an architectural antipattern. If an engineer can trigger an outage by missing a notification on a Friday afternoon, the failure belongs to the system design. Safety must be enforced through code, automated pre-flight gates, and structural IAM constraints.
  • Remove the traps from the floor: Platform maturity is not about asking teams to walk through minefields more carefully. It is about building abstractions so clean that developers cannot accidentally violate least privilege, leave egress unmonitored, or trigger runaway warehouse costs.
  • Documentation as operational self-care: Architectural Decision Records (ADRs) and deterministic runbooks are not bureaucracy—they are the primary defense against cognitive overload, team burnout, and human single points of failure.
  • Compliance by construction: In regulated enterprise environments (HIPAA, SOC 2, HITRUST), compliance cannot be a panicked quarterly audit scramble. It must be an immutable, continuously generated byproduct of platform automation.

Capability Domains

Multi-Account Data Platform Governance

Establishing centralized access controls, automated role management, and multi-layer cost controls across distributed data platforms.

  • Multi-account Snowflake RBAC automation and permission risk categorization
  • Databricks Unity Catalog on GCP
  • Multi-layer cost defense (warehouse policies, resource monitors, and connection reuse)
  • Cross-account change detection with graceful degradation

Secure Connectivity & Data Access

Designing network isolation and controlled data access across AWS and GCP for sensitive workloads.

  • Zero Trust Network Access (ZTNA) and hybrid multi-cloud connectivity
  • Secure database access layers with connection pooling
  • VPN connectivity automation across multiple VPCs

Identity, Access & Compliance Governance

Translating regulatory requirements (HIPAA, SOC2, HITRUST) into automated platform controls and verifiable audit trails.

  • Multi-cloud IAM auditing and automated permission risk categorization
  • GitHub Actions OIDC federation (zero long-lived credentials in CI/CD)
  • Credential rotation without downtime
  • Automated compliance reporting and audit readiness

Standardized Infrastructure as Code

Building reusable, tested, and documented IaC module libraries that enforce architectural guardrails across teams.

  • Multi-cloud OpenTofu and Terraform module libraries
  • Validation and pre-deployment checks
  • State management and backend configuration standards

AI Platform Security & Governance (Current Focus)

Applying platform engineering discipline and Zero Trust security principles to agentic workflows and AI integrations.

  • Zero Trust boundaries and least-privilege constraints for Model Context Protocol (MCP) integrations
  • Operational observability patterns for non-deterministic AI workloads
  • Agent-assisted engineering workflows with deterministic guardrails and pull-request verification

Technical Stack Summary

  • Cloud: AWS and GCP (multi-account)
  • Data Platforms: Snowflake, Databricks, BigQuery
  • Infrastructure as Code: OpenTofu, Terraform, GitHub Actions
  • Languages: Python, TypeScript/Node.js
  • Compliance: HIPAA, SOC2, HITRUST

Current Focus

Platform Engineering at Regulated Scale

Operating platforms in regulated environments requires systems designed for auditability, resilience, and strict access boundaries:

  • Regulated multi-cloud scale: Multi-account data warehouse governance and OpenTofu/Terraform module libraries across AWS and GCP.
  • Safe-by-default tooling: Administrative tooling executes in preview mode by default. Destructive or privileged modifications require explicit elevation, preventing unintended changes.
  • Compliance by design: Automated change detection, least-privilege RBAC, and multi-layer cost controls embedded directly into platform infrastructure.
  • Resilience through graceful degradation: Administrative and audit pipelines tolerate single-account maintenance or transient network partitions without halting broader governance sweeps.
  • Operational documentation: Two-layer documentation discipline combining clear operator runbooks for engineering teams with structured architectural decision records.

Beyond Code

After 25+ years in infrastructure, spanning on-premise data centers to multi-cloud at scale, I have learned that sustainable performance engineering applies to systems AND people.

As a neurodivergent engineer, I approach complex systems with pattern recognition that is both a strength and a responsibility. The same principles that prevent infrastructure burnout (observability, graceful degradation, capacity planning) apply to career sustainability.

The tech industry often celebrates “hustle culture” and endless availability. I have learned that the engineers who last decades (and who ship reliable systems consistently) treat their own capacity as seriously as they treat system capacity. Monitoring your own metrics matters as much as monitoring your infrastructure.

I occasionally write about burnout prevention, neurodivergence in tech, and building careers that last decades, not just sprints.

Want to connect? Find me on GitHub, LinkedIn, or email.