Mario Erazo

Engineering Manager and Solution Architect. Building resilient cloud platforms, leading distributed teams, and engineering workflows for AI agents.

Cologne, Germany

Three lines, one network. Each stop is a topic the line serves; each line ends at the outcome it delivers. The ring around them is how I work, on every line.

Network map Three lines, one for each area of expertise, with four stations for its topics and an endstation plate with its outcome, inside a ring line of nine working principles. Stations and plates link to the sections below. Contextfirst Peerleadership Teamalignment Sharedproduct Simplicity Documentation Measuredoutcomes Qualityassurance Compliance Solution architectureand integrationPlatforms that remain a singlemaintainable product while clientnumbers, external systems, andoperational flows scale. Platformdesign Integrationarchitecture Processmapping Costarchitecture EngineeringleadershipStructured delivery, cleartechnical direction, andpredictable releases acrossdistributed engineering teams. Workflowdesign Platformconsolidation Deliveryorganisation Cross-teamcoordination AI-assistedengineeringCodebases, observability, anddelivery toolchains re-engineeredso AI agents execute tasks acrossthe software lifecycle safely. Agent-readycodebases Runtimeobservability Workflowautomation Teamenablement
Solution architecture and integrationPlatforms that remain a single maintainable product while client numbers, external systems, and operational flows scale.
  1. Platform design
  2. Integration architecture
  3. Process mapping
  4. Cost architecture
Engineering leadershipStructured delivery, clear technical direction, and predictable releases across distributed engineering teams.
  1. Workflow design
  2. Platform consolidation
  3. Delivery organisation
  4. Cross-team coordination
AI-assisted engineeringCodebases, observability, and delivery toolchains re-engineered so AI agents execute tasks across the software lifecycle safely.
  1. Agent-ready codebases
  2. Runtime observability
  3. Workflow automation
  4. Team enablement

Ring line, on every line

  1. Context first
  2. Peer leadership
  3. Team alignment
  4. Shared product
  5. Simplicity
  6. Documentation
  7. Measured outcomes
  8. Quality assurance
  9. Compliance

Expertise

Three lines

Each line is one area of work: four stations for the topics it serves, and at its end the plate with the outcome it delivers.

Line 1

Solution architecture and integration

Platforms that remain a single maintainable product while client numbers, external systems, and operational flows scale.

  1. Platform design

    Cloud infrastructure and container orchestration designed so that client requirements become configurable platform features rather than custom forks.

  2. Integration architecture

    Adaptable integration layers decoupling the core platform from diverse client systems across batch files, synchronous APIs, and event streams under mutual security standards.

    In practiceClient systems using disparate protocols connect securely to one core platform without fragmenting the shared codebase.

  3. Process mapping

    Translating disparate client operational workflows and external data formats into the platform’s core data model to support multi-vendor marketplace flows.

    In practiceComplex external operational workflows and data formats map cleanly into the internal data model without altering core entities.

  4. Cost architecture

    Consolidating isolated client deployments and dedicated application data stores into shared multi-tenant infrastructure alongside leaner compute, caching, and logging stacks.

    In practiceCloud expenditure drops when isolated client environments and dedicated databases are consolidated into shared multi-tenant infrastructure.

Line 2

Engineering leadership

Structured delivery, clear technical direction, and predictable releases across distributed engineering teams.

  1. Workflow design

    End-to-end engineering workflows built from scratch with defined stages and complete visibility across system components and releases.

    In practiceDistributed teams across operations, development, and quality assurance operate as a unified delivery unit once workflow stages and component tracking are established.

  2. Platform consolidation

    Unifying components, delivery processes, and technical standards across the organisation to reduce active services and operational overhead.

    In practiceFragmented platforms become simpler to operate and faster to deploy once repositories, components, and technical standards are unified.

  3. Delivery organisation

    Coordinating operations, development, and quality assurance as one unit, turning commercial requirements and security requests into a sequenced roadmap.

  4. Cross-team coordination

    Sequencing interface contracts and bidirectional sprint dependencies across internal teams and external contractors in large programmes.

    In practiceMulti-team programmes deliver predictably when interface contracts and mutual sprint dependencies are sequenced upfront across all participating groups.

Line 3

AI-assisted engineering

Codebases, observability, and delivery toolchains re-engineered so AI agents execute tasks across the software lifecycle safely.

  1. Agent-ready codebases

    Consolidating dependent repositories for full visibility, enriched with system-architecture manifests and agent instruction rules.

    In practiceChanges that stall on missing context reach production faster once dependent repositories are merged and documented with architecture manifests.

  2. Runtime observability

    Instrumenting application components and logs so AI agents retrieve runtime telemetry for automated verification and debugging.

  3. Workflow automation

    Custom integration tooling connecting version control and issue tracking so agents execute commits, merge requests, ticket transitions, and releases.

    In practiceRoutine delivery overhead shrinks when custom integrations connect version control and issue tracking, enabling agents to automate routine coding workflows.

  4. Team enablement

    Establishing guardrails, specifications, and review gates so engineering teams integrate AI agents into daily delivery safely.

    In practiceBuilt this website from one specification that several AI models implement independently, allowing direct comparison of their output.

Further evidence

A side branch

Other stops the same lines have passed.

  • Fragmented data sources become findable through a unified, federated search architecture.
  • Product catalogues from multiple retailers ingest and index into a shared, multi-tenant search platform.
  • User behaviour events transform into personalised product suggestions through collaborative-filtering recommendation engines.
  • Core microservices run reliably on container orchestration platforms with automated build and deployment pipelines.
  • AWS Certified Solutions Architect – Associate

How I work

Ring line

The ring unrolled: nine stops that every line passes, whichever area the work is in.

  1. Context first

    I gather business and technical context first, weigh trade-offs, consult building engineers on an RFC, and record decisions in an ADR.

  2. Peer leadership

    I lead through technical clarity and structured coordination, supporting engineers while keeping delivery moving and adapting to valid counterarguments.

  3. Team alignment

    I conduct technical interviews, hold regular one-to-ones, and align task delegation with platform requirements and individual engineering growth goals.

  4. Shared product

    I evaluate client requests commercially and architecturally, generalizing viable requirements into core features while isolating custom requests behind configuration flags.

  5. Simplicity

    Every separate service, database instance, and repository adds cognitive load. I treat platform consolidation as a concrete deliverable scheduled alongside features.

  6. Documentation

    System-architecture manifests and agent instructions live inside each repository so humans and AI agents see component impact before modifying code.

  7. Measured outcomes

    I evaluate changes by their observable effect on cloud expenditure, lead time from ticket creation to production, and release frequency.

  8. Quality assurance

    Quality is part of the delivery workflow, not an afterthought. I coordinate QA alongside development so changes are verified across all components before release.

  9. Compliance

    Security and compliance requests enter the backlog alongside commercial features, translated into sequenced work packages so mutual standards build naturally into the roadmap.

Background

The line so far

Five stops on the main line, with a branch line for education.

Since 2017, designing, building, and leading cloud-native commerce and search platforms, progressing from software engineering and technical product ownership to solution architecture and engineering management across distributed teams.

  1. Software Developer Enterprise Digital Commerce Collaborative-filtering recommendation engine and core microservices on cloud and container infrastructure
  2. Technical Product Owner, Search and Recommendation Enterprise Digital Commerce Multi-tenant search SaaS product and federated search delivery for enterprise retail and aviation
  3. Technical Product Owner Enterprise Software and Logistics Fleet, logistics, and contract management software in a multi-vendor scaled agile programme
  4. Solution Architect B2B Travel Retail and Aviation Commerce External integration architecture, domain process mapping, and multi-tenant cloud cost reduction
  5. Engineering Manager B2B Travel Retail and Aviation Commerce Coordinating operations, development, and QA teams; workflow redesign; platform consolidation; AI-assisted engineering

Branch line Education

  1. 2024 M.Sc. Business Management Digital Business Management, completed part-time alongside a full-time role
  2. 2017 M.Sc. Computer Science and Media Machine Learning and Software Development Thesis on ranking search results by user behaviour
  3. 2015 B.Sc. Media Informatics Web Development, Usability, and Accessibility

Languages

  • Spanishnative
  • GermanC2
  • EnglishC2