# Agent Orchestration Projects at AI Tinkerers

> Canonical HTML: https://aitinkerers.org/technologies/agent-orchestration
> Markdown URL: https://aitinkerers.org/technologies/agent-orchestration.md
> Technology record last updated: 2026-04-20T20:37:58Z
> Generated: 2026-09-21T10:50:12Z

Agent orchestration is the command layer that coordinates multiple specialized AI agents to solve complex, multi-step workflows.

Think of orchestration as the conductor for an AI ensemble: it manages state, handles handoffs between models like GPT-4o and Claude 3.5, and ensures logical consistency across long-running loops. Instead of a single prompt hitting a wall, frameworks like LangGraph or CrewAI allow a 'Researcher' agent to browse the web, pass structured data to a 'Writer' agent, and trigger a 'Critic' agent for final validation. This architecture moves beyond simple RAG (Retrieval-Augmented Generation) by implementing cycles, conditional branching, and human-in-the-loop checkpoints to reduce hallucinations and automate production-grade business logic.

- Official technology site: https://www.langchain.com/langgraph
- Public AI Tinkerers demos and talks: 2
- Result page: 1 of 1

## Recent Public Talks and Demos

### [From Local Agent to Emergent Data Scientist](https://nurnberg.aitinkerers.org/talks/rsvp_T0ZyJGZ5BE0)

A local agent gets a stateful Jupyter kernel and a DuckDB database connection instead of fixed tools, giving it a full Python code execution environment. It has to predict outcomes for head and neck cancer patients from cohort data alone, with the current patient's outcome hidden. What I didn't design: the agent trained its own ML models and weighted their outputs based on the patient's individual features, unprompted. Live: I'll show agent traces and behavior in Logfire, walk through the system design, and explain the sandboxed code execution setup.

- Event context: AI Tinkerers Nürnberg: June Meetup (Community Hub) — 2026-06-24 — Nürnberg
- Public talk page: https://nurnberg.aitinkerers.org/talks/rsvp_T0ZyJGZ5BE0

### [Agents require executive function](https://toronto.aitinkerers.org/talks/rsvp_8my42K3UIdc)

I'm building out an open source YAML specification for agent orchestration. Based on a hundred years of human cognitive science, the metaphor is the construction of an artificial Neocortex. Specifically working on improving executive function in agents.

- Event context: AI Tinkerers Toronto - April 2026 - hosted by Shopify — 2026-04-29 — Toronto
- Public talk page: https://toronto.aitinkerers.org/talks/rsvp_8my42K3UIdc

## Related Technologies

- [Artificial Neocortex](https://aitinkerers.org/technologies/artificial-neocortex) ([Markdown](https://aitinkerers.org/technologies/artificial-neocortex.md)) — 1 public demo
- [DuckDB](https://aitinkerers.org/technologies/duckdb) ([Markdown](https://aitinkerers.org/technologies/duckdb.md)) — 9 public demos
- [Jupyter](https://aitinkerers.org/technologies/jupyter) ([Markdown](https://aitinkerers.org/technologies/jupyter.md)) — 8 public demos
- [Jupyter kernel](https://aitinkerers.org/technologies/jupyter-kernel) ([Markdown](https://aitinkerers.org/technologies/jupyter-kernel.md)) — 1 public demo
- [llama](https://aitinkerers.org/technologies/llama) ([Markdown](https://aitinkerers.org/technologies/llama.md)) — 48 public demos
- [LLM](https://aitinkerers.org/technologies/llm) ([Markdown](https://aitinkerers.org/technologies/llm.md)) — 123 public demos
- [Logfire](https://aitinkerers.org/technologies/logfire) ([Markdown](https://aitinkerers.org/technologies/logfire.md)) — 2 public demos
- [Observability](https://aitinkerers.org/technologies/observability) ([Markdown](https://aitinkerers.org/technologies/observability.md)) — 2 public demos
- [Pydantic AI](https://aitinkerers.org/technologies/pydantic-ai) ([Markdown](https://aitinkerers.org/technologies/pydantic-ai.md)) — 7 public demos
- [Python](https://aitinkerers.org/technologies/python) ([Markdown](https://aitinkerers.org/technologies/python.md)) — 662 public demos
- [YAML](https://aitinkerers.org/technologies/yaml) ([Markdown](https://aitinkerers.org/technologies/yaml.md)) — 11 public demos