Roadmap, What's Here and What's Next

This library ships as a living resource. Here's what exists today and what's on deck. Want to move something up the list, or build it with us? See Contribute.

Available now

• Getting started, your first hour with AI, safely.

• The five core guides, client communication, tax research, writing & memos, automation, data cleanup.

• Compliance myths + the Guardrails and cited Regulatory Foundation.

• AI Tax Regulatory Study Guide, a 12-lesson self-study course on the full regulatory stack (§7216, FTC/WISP, AICPA, SSTS, Circular 230, state law, local models), citation-verified.

• Worked examples, tax research, end-to-end and an email before/after.

• Tools & templates, the Redactor redaction tool, the tax-research skill, and document templates (AI policy, §7216 consent, research memo, fact sheet).

Open Lab track (for builders), now live

• Build vs. Buy framework, when to build, when to buy, what to ask vendors.
• Skills & "scripts vs LLM" architecture, incl. the staging-table write-safety pattern.
• Workpaper automation, the most-requested build.
• Local AI models, the strongest safe lane: run a model on your own hardware so nothing leaves your walls, with the hardware tiers, model picks, and the one cloud component that breaks the §7216 benefit.

Hands-on interactive labs, now live

Self-checking, coached walkthroughs that run on your own model and drill "code computes, AI drafts, you decide":

• Bookkeeping pipeline, statements → trial balance → bank rec.
• 1040 review, catch what an AI review misses.
• On deck: month-end close, document processing, and more, same engine.

On deck (shaped by what you ask for)

Drawn straight from the Lab's calls, what members want next:

• Open Lab: MCP connectors & dashboards, rolling your own, multi-entity close, write-safety at scale.
• Open Lab: Cost & model strategy, plan tiers, Sonnet-vs-Opus right-sizing, the extended-thinking token trap.
• Open Lab: Advanced data security, local LLMs, zero-retention BYO-LLM, multi-window PII validation.
• Grounding tax authority, OpenAccountants/verified rulesets, Blue J / TaxGPT (still verify every cite).
• Client-deliverables stack, Gamma / Loom decks, voice intake.
• Engagement-letter §7216 AI clause + a build-vs-buy-style decision aid.
• Advisory / CFO / FP&A, bookkeeping automation, team rollout, and more worked examples.

Help shape it

• Vote with your requests, the most-asked topics get built first.
• Multistate, the regulatory material is Georgia-anchored; other states welcome as contributions.
• Found a gap or an error? That's a roadmap item too, tell us.

Part of the AI Lab for Accountants library. The roadmap moves when members push it.