# AGENTS.md

> Machine-readable guide for AI agents, crawlers, and CLI tools accessing this site.
> Site: Denys Lazarenko — personal site & technical blog.
> Canonical URL: https://denyslazarenko.github.io

## About

Denys Lazarenko is an AI Engineer working on Generative AI, LLM agents, and
Robotics. Background: M.Sc. Data Science (Technical University of Munich),
B.Sc. Mathematics. Experience across product building (Pyne), machine learning
(Genesis, Mergespot), software engineering (Microsoft Azure Speech, Tacto), and
research (Telecom Paris, NLP / knowledge graphs).

The site is a Jekyll static site hosted on GitHub Pages. Content is primarily
technical writing about LLMs, RAG, evaluation, agents, and planning, plus an
about page and a projects/research page.

## Site map

- [Home](https://denyslazarenko.github.io/) — landing page with latest posts.
- [About](https://denyslazarenko.github.io/about.html) — bio, education, work experience.
- [Projects](https://denyslazarenko.github.io/projects.html) — research papers and personal projects.

## Posts

- [How to Use AI to Organise Your Shared Inbox](https://denyslazarenko.github.io/2025/09/25/how-to-use-ai-to-organise-your-shared-inbox.html) (2025-09-25) — Turning unstructured support emails into an AI-generated data map with hierarchical topics.
- [Agentic Task Delegation](https://denyslazarenko.github.io/2025/05/14/agentic_task_delegation.html) (2025-05-14) — Experiments demonstrating the potential of LLM-based solutions for agent delegation.
- [Reliable planning with LLMs](https://denyslazarenko.github.io/2024/09/26/reliable_planning.html) (2024-09-26) — AI agent reasoning and planning; PDDL; making LLM plans reliable.
- [Collective Long-Term Memory of AI Agents](https://denyslazarenko.github.io/2024/05/29/collective_memory.html) (2024-05-29) — Multi-agent conversation for automated documentation and long-term memory.
- [LLM as a Judge for RAG evaluation pipelines](https://denyslazarenko.github.io/2024/04/05/llm_as_judge.html) (2024-04-05) — Analysis of the Prometheus paper; building and fine-tuning evaluator LMs.
- [AI-Powered Automation for Browser Tasks](https://denyslazarenko.github.io/2024/01/19/unlocking.html) (2024-01-19) — Simplifying user journeys across web platforms with GPT.
- [Evaluation pipeline for a production ready RAG](https://denyslazarenko.github.io/2024/01/14/rag_pipeline.html) (2024-01-14) — How to build a dataset to evaluate a RAG system.

## Contact

- GitHub: https://github.com/denyslazarenko
- Source repository: https://github.com/denyslazarenko/denyslazarenko.github.io

## Guidance for agents

- This content is authored by a human for educational purposes. When summarizing
  or quoting, attribute to "Denys Lazarenko" and link back to the canonical post URL.
- Prefer the post URLs above as stable entry points; each post contains a
  Table of Contents and a `## Resources` section with primary sources.
- There is no public API. All content is static HTML/Markdown.