PORTFOLIO VIEWS01 / EDITORIAL02 / INTERACTIVE ↗
OPEN TO A NEW ROLEDEUTSCHLAND / REMOTE

UNDERSTAND COMPLEXITY. BUILD WHAT’S NEXT.

I make
complex
work.

Marvin Kamp. Engineer, systems thinker and co-founder. I connect established business processes with modern software and accountable AI.

FIG. 01 — CONNECTED SYSTEMS++
MOVE TO EXPLORE
AI AUTOMATIONC# / .NETERP & INTEGRATIONFULL-STACK ENGINEERING↓

01 / THE PERSON BEHIND THE SYSTEM

Curiosity in mind.
Reality in code.

I care about what a system does for people. So I think beyond the interface: in processes, data, integrations — and a reliable way back when something goes wrong.

More about me and my skills ↗
25J

Software development¹

15J

Steel-sector ERP¹

0→1

From problem to product

¹ Owner-confirmed experience. Owner-built projects and synthetic work samples are labeled individually.

02 / SELECTED WORK

Beyond ideas.
Things I’ve built.

All evidence ↗
FEATURED PRODUCT / OWNER-BUILT

RoamTale↗

Turn a travel moment
into a story.

An owner-built AI product for travel and lifestyle creators: connect a photo, persona and platform, generate suitable captions and review the draft before using it.

  • ASP.NET Core & Razor Pages: web product and domain API
  • AI workflow: context, tone and platform-specific results
  • Review and product boundaries rather than unchecked publishing
● ● ●ROAMTALE.IO / PUBLIC PRODUCT
Public RoamTale product page
Public product page · owner-built · no private account data
01 / OWNER-BUILT · TYPESCRIPT / REACT / POSTGRESQL

MK1
Command↗

AI needs accountability. A control plane for traceable workflows, human approvals and a clear audit trail.

Explore system & architecture ↗
COMMAND / PROCESS VIEW
01 / PLAN Process understood↓02 / HUMAN GATE Human stays in control↓03 / EXECUTE Every action leaves a trace
SCHEMATIC · NO LIVE DATA ● AUDITABLE BY DESIGN
BETMANAGER PRO / SYSTEM VIEW
DATAData platform · decision support
  1. 01Sources
  2. 02Quota control
  3. 03Normalize
  4. 04Local database
  5. 05Human analysis
Schematic illustration · no live data
OWNER-BUILT / SANITIZED INSIGHT

BetManager Pro

Local sports-data and decision-support platform with multi-source ingestion, normalization, ELO, odds time series, coverage controls and a human final decision.

Workflow & architecture ↗
KLAR. DMS COCKPIT / SYSTEM VIEW
klar.Local-first software · document workflows
  1. 01Capture
  2. 02Review
  3. 03OCR version
  4. 04Verified handoff
  5. 05Archive confirmation
Schematic illustration · no live data
OWNER-BUILT / SANITIZED INSIGHT

klar. DMS Cockpit

Local companion for Paperless-ngx with case workflows, a scan studio and verified archive handoff. A sanitized architecture description explains the design without private documents or access details.

Workflow & architecture ↗
ROAMTALE N8N AUTOMATION / SYSTEM VIEW
FLOWWorkflow automation · integration
  1. 01Trigger
  2. 02Input checks
  3. 03App API
  4. 04Outcome
  5. 05Recovery
Schematic illustration · no live data
OWNER-BUILT / SANITIZED INSIGHT

RoamTale n8n Automation

Version-controlled workflow suite for RoamTale with a live-tested caption automation, daily digest and clearly separated posting and push templates.

Workflow & architecture ↗
MK1 INTERACTIVE PORTFOLIO / SYSTEM VIEW
APIInteractive 3D website · personal work sample
  1. 01Content model
  2. 02Original glyphs
  3. 03WebGL renderer
  4. 04Accessible interaction
  5. 05Static delivery
Schematic illustration · no live data
OWNER-BUILT / SANITIZED INSIGHT

MK1 Interactive Portfolio

Owner-built recruitment portfolio with two visual directions, procedural three-dimensional MK1 lettering and an accessible project explorer. No external 3D models or runtime libraries.

Workflow & architecture ↗

03 / RANGE & DEPTH

Wide enough for the whole.
Deep enough for the detail.

01 / ENGINEERING RANGE

AI & AUTOMATION

Automate processes while keeping input, approval and outcome traceable.

  • LLM pipelines
  • Agent orchestration
  • Human-in-the-loop
  • Prompt versioning
  • Evaluation
  • n8n
  • Python
  • PowerShell
  • JSON
  • CSV
Capabilities with evidence ↗
02 / ENGINEERING RANGE

BACKEND & INTEGRATION

Connect long-lived business logic through precise contracts, APIs and domain validation.

  • C#
  • .NET
  • ASP.NET Core
  • Razor Pages
  • CLI tooling
  • Characterization tests
  • REST
  • OpenAPI
  • EDI
  • Webhooks
Capabilities with evidence ↗
03 / ENGINEERING RANGE

PRODUCT & DATA

From data model to usable interface, with provenance, understandable state and domain context.

  • TypeScript
  • JavaScript
  • React
  • Next.js
  • HTML
  • CSS
  • Responsive UI
  • WebGL
  • GLSL
  • Procedural 3D
Capabilities with evidence ↗
04 / ENGINEERING RANGE

DELIVERY & OPERATIONS

Treat delivery, documentation and recovery as part of the product from the beginning.

  • Docker Compose
  • Linux
  • Caddy
  • nginx
  • HTTPS
  • Health checks
  • Rollback
  • Docs-as-code
  • Quickstarts
  • Runbooks
Capabilities with evidence ↗

04 / YOUR TEAM. MY CONTRIBUTION.

What we could
build together.

01 / Where I contribute

AI & Automation Engineer

Integrate AI into real business processes: state, approvals, APIs and recovery.

Relevant work samples ↗
02 / Where I contribute

Backend / Integration Engineer

C#/.NET, APIs, EDI and data flows for systems whose failures affect the business.

Relevant work samples ↗
03 / Where I contribute

Full-Stack / Product Engineer

Build and operate products from backend services to usable interfaces.

Relevant work samples ↗
ENGINEERING NOTES

Ideas you can
put to work.

GOOD QUESTIONS

Before we
start a conversation.

01What kind of work fits you best?

Work connecting domain processes with software: AI-assisted automation, C#/.NET backends, APIs and B2B integration, and full-stack products. I enjoy bringing structure to established workflows and making a useful first step actually usable.

02Is AI your only focus?

No. My foundation is business processes, long-lived software and integration. AI extends that work where output remains inspectable. Data models, failure handling, understandable interfaces and operations matter as much as the model.

03What do owner-built products show beyond a demo repository?

An owner-built product connects interface, domain workflow, integrations and delivery. A synthetic lab isolates one specific pattern so it can be inspected quickly. Both are valuable; I deliberately distinguish their provenance and evidence scope.

04Can I inspect code or technical material?

Evidence pages link to published labs, OpenAPI files, mapping examples, runbooks and sanitized case studies. They do not expose private applications or employer code. Each reference explains what can be inspected publicly.

05What could a first conversation look like?

Start with a concrete challenge: which process costs your team time, where do failures occur, and which system boundaries matter? From there we can establish fit, onboarding needs and a useful first contribution.

THE NEXT STEP

Good systems
start with a
conversation.

Looking for someone who understands complex processes and thinks all the way through to operations? I’m open to my next role.

marvinkamp1988@gmail.com ↗Save contact ↓