SebastianHeitmann

Output is cheap now. Outcomes are not.

AI made producing software easy. Deciding what to build, judging what's good, and carrying it into production didn't get easier. Developer, project manager, consultant: I've been all three. Today they're one role. I build AI-powered products and processes and own them from first prototype to production.

Send a Message→Direct to my inbox. No funnel. I read every message.
Previously at
Jung von MattsynvertGranny & SmithOFFIS
What I do
01

Software Development

  • App Development
  • Fullstack Development
  • API & Backend Systems
  • Cloud Infrastructure
02

AI & Automation

  • Agentic Workflows
  • Prompt & Context Engineering
  • AI-assisted Development
  • Human-in-the-Loop Systems
03

Architecture & Compliance

  • System Architecture
  • Domain-Driven Design
  • GDPR & Data Protection
  • AI Trust Boundaries
04

Concept, Planning & Management

  • Technical Leadership
  • Requirements Shaping
  • Design Sprints
  • Rapid Prototyping
When to reach out
  • Delivery has slowed, and the team sees a different cause in every direction
  • The architecture got you here, but it will not carry the next stage
  • A technical decision now will shape cost, speed, or team structure for years
  • A migration, rebuild, or reorg needs someone who has already done it
Results
Gap Closed54% project risk identified and addressed

Identified a 54% projected failure risk through an initial assessment, then restructured the project management and changed the execution conditions to create a more viable foundation for the initiatives that followed.

Execution Risk
9+Contributors

Led technical project management for migrating a legacy SAP BW analytics environment to AWS, coordinating across architecture, data engineering, and delivery to build a scalable reporting foundation.

Cloud Migration
7countries rolled out

Stabilized the platform and content pipeline ahead of launch, enabling a reliable rollout across seven countries from day one.

Technical Delivery
Stalled → Stableintranet relaunch

Aligned concept, planning, budget approval, and cross-team execution for the relaunch of an internal platform — turning a stalled initiative into a stable delivery.

Product Leadership
+2points in satisfaction

Introduced a decentralized IT support model to improve internal service quality, contributing to a measurable two-point increase in satisfaction across the organization.

Operational Improvement
Engagement
AdviseHourly

Billed by the hour. Architecture reviews, hiring decisions, vendor evaluations, and second opinions before a high-impact commitment.

Technical project management→
DeliverPer Cycle

Fixed price per delivery cycle. A focused engagement around a defined outcome — an MVP, a migration, a rebuild, or a delivery reset.

AI-powered products & processes→Web development for local businesses→
PartnerMonthly Retainer

Ongoing monthly retainer. Senior technical partnership across architecture, delivery, and execution continuity.

Writing
All articles9→
Control Isn't Free: Why Startups Copy Enterprise Governance They Can't Afford
Leadership & ManagementJun 11, 2026

Control Isn't Free: Why Startups Copy Enterprise Governance They Can't Afford

Small companies are starting to see bottom-up AI automation before they have mature governance around it. The reflex is to copy enterprise controls: approval gates, tool lists, review processes, and policy language. But control is operational overhead, not free insurance. This article argues that startups and SMEs need to price control against risk, protect genuinely fatal zones, and avoid destroying the adapted capacity their fastest employees are already building.

AI AdoptionGovernanceLeadershipStartupsSME LeadershipAutomationExperimentation
Why Buying AI Tools And Training Employees Is Not AI Adoption
Leadership & ManagementMay 23, 2026

Why Buying AI Tools And Training Employees Is Not AI Adoption

Many companies will buy AI tools, train employees, and still fail to build AI capability. The problem is not weak technology. It is a shallow adoption model that treats AI like another software rollout instead of a change in how judgment, delegation, experimentation, and infrastructure move through the organisation. This article explains why training alone stops too early, why AI moves people from execution toward steering, why apprenticeship and management judgment matter, and which operating-model question executives should ask before adoption becomes theatre.

AI AdoptionLeadershipOperating ModelAI StrategyDigitalisationOrganisational Change
Stop Asking If SaaS Dies. Ask Which.
Technical DecisionsMay 1, 2026

Stop Asking If SaaS Dies. Ask Which.

An M&A expert bet that companies would stop buying SaaS because AI made building in-house cheap enough. That theory does not survive contact with organisational debt. AI made parts of software cheaper, but the expensive part was never only the code — it was the structure that dictates processes, the processes that dictate tooling, and the ownership questions that make every adoption curve slower than the demo suggests. The useful question is not whether SaaS survives AI. It is which SaaS survives AI.

SaaSAIBuild vs BuySoftware IndustryM&AOrganisational Debt

Start a Conversation

Tell me what you are working through. I will get back to you within one business day with a candid view on whether and how I can help.