Shmuel Koltov

A little about me

I’m Shmuel Koltov, a Support Operations Specialist at MyHeritage, where I focus on AI and automation. In practice that means building N8N workflows, AppSheet applications, and the information flows that connect platforms and make data move usefully between them. I also work with local LLM stacks, design knowledge systems, and spend a lot of time thinking about how the pieces fit together rather than how any single piece works in isolation.

My academic background is in comparative religion and modern Middle Eastern studies, which might sound like it has absolutely nothing to do with what I’m doing today. That’s not wrong if you look at it without considering it deeper, but while the subject was religion, my studies had at least as much to do with how you learn to read complex systems and resist the pull of simple explanations as it does with religion. That particular habit of mind turns out to be useful when you need to think in systems or work with dogmatic thinking.

The work itself is mostly about connections: how information moves between systems, where it gets lost or distorted, and what needs to be in place for AI to be genuinely useful rather than just present. A lot of that involves people as much as platforms. Understanding how someone actually works, where they’re stuck, and what would make the difference is just as important as knowing which tool to reach for. I build N8N workflows, AppSheet applications, and knowledge structures. I use AI to do things that would otherwise require skills I haven’t developed. And I’ve learned enough about security, data management, and system design to know whether what I’ve built is actually trustworthy. The coding is a small part of it. The systems thinking is most of it.

I came to all of this from a place of genuine confusion, not knowing what AI was actually for or where I fit in it. I see that same confusion in a lot of people around me, and some of the hesitation that comes with it. This site is for people in that position. Not developers, not people who’ve already figured it out. People who are curious, maybe uncertain, and wondering whether there’s a path from where they are to somewhere more capable. There is. I’ve been walking it, and I write about what I find.

That means sharing what I’m building, what I’m learning, where I’ve gotten things wrong, and what I think about the tools and ideas I’m working with. It’s not instruction. It’s documentation from someone in the middle of it.

---

What I work with

N8N · AppSheet · Claude · Gemini · Google Workspace · Notion · Atlassian · Jira · Confluence · LM Studio · API · JSON · Markdown · Knowledge Graphs · Neo4j · Cypher · RAG · AI Agents · Agentic AI · Workflow Automation · Local LLMs · Prompt Engineering

---

What I’m studying and interested in

AI · Generative AI · LLM · Data Management · Information Flows · Knowledge Management · Knowledge Engineering · Systems Thinking · Vector Databases · Graph Databases · Relational Databases · Vector Indexing · AI Memory Engineering · AI Context Engineering · AI Integrations · Mistral · RAG · AI Agents · Prompt Engineering · Local LLMs

---

A note on how this site is made

I work with AI as a writing partner and collaborator. The thinking, the positions, and the editorial judgment on this site are mine. The prose is produced with AI assistance. I’m transparent about this because it’s consistent with everything I argue here: AI extends what a person can do. This site is an example of that, not a contradiction of it.