Orchestrate Agentic AI

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

• Time: 30min
• Presenter: Calvin
• Event: https://2026.allthingsai.org

Abstract

Orchestrate Agentic AI: Context, Checklists, and No-Miss Reviews

The talk description/pitch:

Have you ever run a 20-page contract through an AI and watched it miss the clause that actually mattered? You point it out and the model replies, “Oh, you’re right.” You are not alone, which raises a hard question: how can AI be trusted in high-stakes environments?

AI practitioners are under pressure to deploy AI on contracts, policies, and code reviews without increasing risk.

Yet most teams > still rely on chat-style prompts and generic tools that miss key > clauses, skip steps, and make it hard to prove what was actually > reviewed. New tool frameworks and MCP-style integrations are > multiplying calls and windows, but not reliability. The result is > stalled negotiations, audit anxiety, governance gaps, and persistent > doubts about whether AI is safe to trust.

Join Calvin Hendryx-Parker (AWS Hero and CTO of Six Feet Up, the premier AI and Python agency in the U.S.) for a practical look at orchestrating agentic AI systems. Learn how to turn generative AI into a dependable reviewer for contracts, policies, audits, and software development workflows such as code review and test planning.

What You’ll Learn

• Context Engineering Tactics: How to chunk documents and tasks, frame targeted prompts, and apply checklists so each sub-task has clear acceptance criteria and traceability.
• Agentic Workflow Strategies: How to turn a vague “review this” into a structured series of sub-jobs, including clause-level analysis, routing, and validation so nothing is silently skipped.
• No-Miss Review

Demo: A live, clause-level walkthrough of an AI-driven contract review that shows task graphs, specialized agents, routing, and validation, plus direct parallels for engineering workflows. Walk away with an actionable playbook to make AI reliable, auditable, and ready for high-stakes enterprise use.

Build this Presentation

Build dependencies

• nvm
• pandoc
• bsdtar (part of the libarchive-tools package on Ubuntu)

Make sure that your working directory is set up to use the LTS version of node before you begin. Then can just build your presentation with make.

$ nvm install --lts
$ make start