Lukas Dabkowski
Menu

Professional evidence

Capabilities

Six connected capability areas grounded in selected professional work. Each section links the methods and technologies to specific evidence while keeping implementation, assessment, guidance, and oversight distinct.

01

Artificial Intelligence & Agentic Systems

Enterprise AI modernization, multi-agent orchestration, and human-gated Agentic workflows with explicit safety and delivery boundaries.

How I apply it

I separate model reasoning, orchestration, tools, validation, approval, and state changes so that each boundary can be reviewed independently.

Related technologies

  • Amazon Bedrock
  • Amazon Nova Pro
  • Python
  • Playwright
  • Salesforce and Certinia
  • AWS CDK
  • Terraform

Evidence from selected work

  • Modernized a deployed internal AI model layer on AWS with guardrails, access controls, and Terraform-managed infrastructure.
  • Implemented a local-first assistant with authenticated browser and PSA integration, explicit approval, draft-only writes, and verification.
  • Designed and implemented an AWS multi-agent proof of concept with specialist tools, structured outputs, infrastructure code, logging, and tests.

02

Cloud & Platform Architecture

AWS architecture plus Azure and Google Cloud assessment experience across target states, platforms, identity, connectivity, governance, and operations.

How I apply it

I connect workload needs with platform services, identity, networking, governance, resilience, and operating responsibility before selecting a target.

Related technologies

  • AWS
  • Azure assessment
  • Google Cloud
  • Amazon EKS
  • Google Kubernetes Engine
  • Serverless services
  • Terraform

Evidence from selected work

  • Assessed an Azure-centered SaaS platform and defined a conceptual AWS target without implying migration execution.
  • Designed AWS recovery and managed-database architectures with explicit engineering, operations, and responsibility boundaries.
  • Implemented bounded AWS infrastructure and controls in selected AI and policy-engineering workstreams.

03

Security & Compliance Engineering

Policy-as-code, regulated controls, cloud findings analysis, least privilege, exception handling, qualification evidence, and guarded automation.

How I apply it

I translate requirements into inspectable controls, validations, exceptions, approvals, and evidence without implying universal compliance ownership.

Related technologies

  • OPAL
  • LQL
  • Terraform
  • CSPM
  • IAM
  • DuckDB
  • Parquet
  • AI guardrails

Evidence from selected work

  • Implemented classification-driven policies and Terraform security modules with differentiated regulated and non-regulated control profiles.
  • Built a local-first firewall evidence pipeline whose proposed cloud changes require human approval and verification.
  • Implemented guardrails, narrow tool contracts, unsafe-state rejection, and access boundaries in defined AI workstreams.

04

Migration & Modernization

Current-state assessment, target design, dependency mapping, migration strategy, implementation guidance, validation, transition, and rollback planning.

How I apply it

I distinguish readiness and target-state decisions from delivery, then define responsibilities, acceptance, operational transition, and rollback before execution.

Related technologies

  • AWS Migration Hub
  • AWS Application Discovery Service
  • Oracle Database
  • Amazon RDS for PostgreSQL
  • AWS DMS
  • Amazon Nova Pro

Evidence from selected work

  • Completed an Azure-to-AWS assessment with dependencies, target-state options, governance, risks, and migration planning.
  • Designed Oracle-to-managed-PostgreSQL migration architecture and enabled the teams responsible for execution.
  • Migrated a deployed enterprise AI model layer while preserving the surrounding application and its access boundary.

05

Resilience, Data & Enterprise Integration

Recovery, managed data, enterprise application integration, workflow state, messaging boundaries, and failure handling designed as one operating concern.

How I apply it

I map data ownership, system boundaries, failure modes, dependencies, recovery actions, and operational handoffs before choosing integration patterns.

Related technologies

  • Amazon RDS for PostgreSQL
  • AWS DMS
  • Amazon Route 53
  • AWS Backup
  • Salesforce REST API
  • Certinia PSA
  • DuckDB
  • Parquet

Evidence from selected work

  • Designed multi-region recovery across runtime, state, traffic, backup, monitoring, failover, failback, and rollback concerns.
  • Designed managed PostgreSQL migration with compatibility, availability, backup, monitoring, secure access, and operating responsibilities.
  • Implemented browser-to-PSA integration with live catalog validation, preserved state, narrow writes, and post-write verification.

06

Architecture Delivery & Evidence

Decisions, assessments, architecture views, implementation material, policy specifications, workshops, operational guidance, and audit or qualification evidence.

How I apply it

I choose artifacts for the decision and audience, connect them to evidence, and record assumptions, boundaries, responsibilities, and unresolved risks.

Related technologies

  • HLD and LLD
  • Architecture decisions
  • Infrastructure as Code
  • Policy as Code
  • Validation evidence
  • Runbooks
  • Technical workshops

Evidence from selected work

  • Produced current-state assessments, target architectures, decisions, migration guidance, and responsibility models.
  • Created policy specifications, infrastructure code, validation, tests, qualification material, and auditable workflow evidence where supported.
  • Delivered technical workshops, runbooks, implementation guidance, and enablement while keeping execution ownership explicit.