iksmgtConsulting DEEN Request a project

IT consulting · Cologne · Remote-first

Process automation for small businesses — with and without AI

Process automation (n8n) · AI integration (LLM, Claude API) · DevOps & self-hosted infrastructure — GDPR-compliant, on-premise or in your cloud

Request a project View services
GDPR-compliantOn-premise or cloudNo vendor lock-in
SOURCES AUTOMATION OUTCOME ERP / IMS E-MAIL / PDF DMS / API n8n WORKFLOW ENGINE AI MODULE LLM · OPTIONAL REPORTING TARGET SYSTEMS AUTOMATED END-TO-END
Fig. 0 — System sketch, schematic
Service index

Services

Sheet 01 / 05 · Rev. C

Three fields, one principle: process first, then tooling. AI joins in where it measurably helps — not the other way around.

POS. S-01

Process Automation (without AI)

  • Workflow automation with n8n
  • ERP, e-mail, DMS & API integration
  • Data integration between systems
  • Reporting pipelines
POS. S-02

AI Automation

  • LLM integration — Claude API, local models
  • RAG pipelines
  • Multi-agent systems
  • LLM-powered document & invoice processing
POS. S-03

DevOps Consulting & Infrastructure

  • Proxmox, Docker, CI/CD
  • Mesh VPN & backup strategies
  • Architecture, setup & operation of self-hosted environments
Standard solutions

Solutions

Sheet 02 / 05 · Rev. C

Recurring back-office problems, solutions engineered once and adapted to your business: proven open-source building blocks — self-hosted, GDPR-compliant, no vendor lock-in.

LSG-01

Digital Document Intake — pilot in 2 weeks

n8n · Paperless-ngx · OCR · AI
INTAKE PIPELINE E-MAIL IMAP INBOX SCANNER SCAN FOLDER FAX FAX-TO-DIGITAL WEB FORM INSTEAD OF PDF n8n · PAPERLESS — SELF-HOSTED OCR TEXT RECOGNITION RULE PARSER DATE · IBAN · AMOUNT AI EXTRACTION TYPE · FIELDS · JSON GDPR-COMPLIANT · DE/EU ARCHIVE FULL-TEXT SEARCH REVIEW LIST WHEN UNSURE HANDOVER DATEV · PROPERTY · ORDER ALL CHANNELS → ONE ARCHIVE
Fig. 1 — Schematic view

Mail, e-mail, scans and faxes land on the desk every day — and get sorted by hand. A digital intake workflow bundles all channels, automatically reads sender, date, amounts and document type, and files every document sorted and full-text searchable. Anything unclear lands on a review list instead of in the wrong folder.

30–60 min of sorting work per person per day eliminated

More about this solution →

Case studies

Selected Projects

Sheet 03 / 05 · Rev. C

Real projects, described anonymized — technology, structure and tooling only. Three examples following the same pattern: problem, architecture, result.

REF-01

Directory platform with a data pipeline from official registries — own product, live

Django · PostgreSQL · HTMX · SEO
DATA PIPELINE REGISTRIES HTML · SCRAPER REGISTRIES PDF · PARSER TAXONOMY CATEGORIES PIPELINE RAW · STAGING · PROD SEO PAGES CATEGORY × PLACE OPS BACKEND n8n · REQUESTS PROFILE PAGES ~870 ENTRIES IDEMPOTENT · APPROVAL GATES OFFICIAL SOURCES — ONE INDEX
Fig. 2 — Schematic view
Problem
The official registry data of a regulated profession is scattered across dozens of registry bodies — HTML lists and PDF registers, with no central search.
Architecture
Server-rendered Django with HTMX and Tailwind on PostgreSQL. Idempotent scrapers load the registries (HTML and PDF) into a three-stage pipeline — raw, staging, production — with manual approval gates between stages. SEO landing pages per category, location and combination; redirect middleware (301/410) keeps the URL structure clean; contact requests flow via n8n into an ops backend (Fig. 2).
Result
A live directory with around 870 profiles from official sources; the index grows in a controlled way, registry by registry. ~870 profiles from official sources
REF-02

Agent workflow: from photo to marketplace listing — internal tool

LLM · Vision · APIs
AGENT WORKFLOW — WITH APPROVAL GATE INPUT PHOTOS + NOTE AGENT WORKFLOW VISION LLM READS THE PHOTOS SCHEMA ATTRIBUTES MARKETPLACE API · LIVE PHOTOS DRAFT IMAGES MATCHING FETCH PUBLISHED ONLY AFTER APPROVAL
Fig. 3 — Schematic view
Problem
Creating a listing on a major online marketplace takes many manual steps: prepare photos, research category and required attributes, write copy, publish.
Architecture
Agent-driven workflow: script-based image prep (background removal, normalizing), a vision model reads the photos and grounds the listing copy in what is actually visible, category and attribute schemas come live from the marketplace API — nothing is published without explicit approval (Fig. 3).
Result
Photos plus a short note become a reviewable listing draft; only approved content gets published — no invented product claims. publishing only after approval
REF-03

Multi-agent pipeline for market research — own product

Python · LLM agents · Scheduling
RESEARCH PIPELINE FILINGS MANDATORY REPORTS TRANSACTIONS INSIDER MARKET SCREENS PRICE · VOLUME SCAN + FILTER ENRICHMENT TWICE A WEEK ADVERSARIAL REVIEW AGENT THESIS AGENT COUNTER-THESIS AGENT RISK REPORT 3–5 IDEAS HUNDREDS OF CANDIDATES → 3–5 IDEAS
Fig. 4 — Schematic view
Problem
Public capital-market signals — mandatory filings, insider transactions, screens — are too numerous to review manually; and single LLM answers are too uncritical to be trusted.
Architecture
A pipeline running twice a week: scan of multiple data sources, enrichment and rule-based filtering, then adversarial review by several agents arguing opposing positions. The output is a report with thesis, catalyst and invalidation level per idea, rendered as a static report site (Fig. 4).
Result
3–5 vetted ideas per run instead of hundreds of raw hits; every idea documents what would prove it wrong. 3–5 vetted ideas per run

See all projects →

Person & approach

About me

Sheet 04 / 05 · Rev. C

I'm Igor Stolyarevskiy, an IT consultant based in Cologne — with a master's degree in IT and more than ten years of professional experience. I design, build and operate automation and self-hosted infrastructure for small and mid-sized companies — from the first process sketch to day-to-day operations. I know the starting point from my own entrepreneurial experience: organically grown systems, manual processes and the question of what automation actually delivers.

I favor self-hosted solutions and open standards — systems you can understand, control and keep operating yourself. AI comes in where it measurably helps.

Inquiry

Contact

Sheet 05 / 05 · Rev. C

The shortest path is an e-mail — two sentences about your process are enough. No cookies, no newsletter.

iks.reachout@gmail.com
Request a project

Reply usually within one business day

Inquiry

Request a project

Briefly describe your process — I usually reply within one business day.

Encrypted transfer to our own infrastructure — never shared with third parties.