FOSSIL RECORD

Fossil Record is a Claude-powered decision-state preservation protocol built on the Snoddy Method. Fossil Record preserves system state, extracting load-bearing joints, confirmed decisions, open variables, collapse risks, and session-specific behavioral guardrails from raw transcripts and compressing them into a structured re-entry prompt with 94 - 97% token reduction. Validated under live platform outage conditions. Peer reviewed across three independent AI nodes.

Fossil Record Pro - AI Session Continuity Manager

AI@USC Workshop Proposal — Jeff Snoddy

01 Title

Clear, engaging title for your workshop

When AI Forgets: How to Build a Workflow That Remembers

02 Abstract

250-word summary including learning outcomes and prerequisites

Every AI session starts from zero. The moment you close a conversation, the context, decisions, and hard-won logic you built with your AI collaborator disappears. For professionals working on complex, multi-session projects — grant proposals, research, program design, strategic planning — this is not a minor inconvenience. It is a structural failure that erodes trust, wastes time, and forces constant reconstruction of work that should compound.

This workshop introduces Fossil Record, a methodology and tool built by Jeff Snoddy — emergency manager, DSW candidate at USC Dworak-Peck, and patent-pending inventor of the Snoddy Method — to solve exactly this problem. Fossil Record captures the logic, decisions, and governance structure of an AI session and reconstitutes it in the next session with full fidelity. It does not summarize. It does not retrieve. It distills and restores the structural logic that makes complex work possible across time.

Participants will work hands-on with a live Fossil Record session, experience the difference between a cold-start AI and a context-loaded AI, and leave with a repeatable workflow they can apply to their own professional or academic projects the same day.

No technical background is required. Participants should bring a laptop and an active Claude, ChatGPT, or Gemini account. The methodology is platform-agnostic.

By the end of this session, participants will understand why AI memory fails, how structural distillation differs from summarization, and how to build a personal AI workflow that holds its logic across sessions.

— words

03 Learning Objectives

2–3 specific outcomes attendees will gain

Participants will be able to explain why AI session discontinuity occurs and identify at least two professional contexts where it creates meaningful risk to their work.

Participants will demonstrate the Fossil Record capture-and-reconstitution workflow by completing a live extraction and re-entry sequence during the session.

Participants will leave with a personal Fossil Record template configured for their own project or research context, ready to use in their next AI session.

04 Target Audience

Experience level and maximum capacity

Experience Level

Beginner to Intermediate

Max Capacity

25 participants

Prerequisites

Laptop + active AI account (Claude, ChatGPT, or Gemini). No coding required.

Who Should Attend

Graduate students, researchers, faculty, and practitioners who use AI in professional or academic work.

05 Detailed Agenda

60-minute session with time allocations

Time	Segment	Description
0:00–0:05	Opening	The problem stated plainly: AI forgets. What that costs in real work.
0:05–0:15	Live Demo — Cold Start vs. Context-Loaded	Side-by-side comparison showing the same AI session with and without Fossil Record. Participants observe the difference before building anything.
0:15–0:25	Concept Instruction	Structural distillation explained. What Fossil Record captures, what it does not, and why this is different from copy-pasting a summary. Equity framing introduced: who loses most when AI workflows fail.
0:25–0:45	Hands-On Build	Participants apply the Fossil Record template to a real project they brought. Facilitator and peer helpers circulate. Participants complete a capture and re-entry sequence with their own content.
0:45–0:55	Debrief and Discussion	What held? What surprised you? Where does this apply in your work? Sovereignty angle surfaced: who owns the logic your AI sessions produce?
0:55–1:00	Close and Next Steps	Where to access Fossil Record. How to build this into a regular workflow. Open questions taken — deeper architecture discussion available offline.

— All six required fields complete. Ready for submission. —

Fossil Record (Snoddy Method v5.0) · Generated March 10, 2026