YOUR ANALYTICS CAPABILITY, AI-USABLE
Make your analytics knowledge AI-ready.
Propperly reconstructs what your organization currently knows from the work your team already has — what’s current, what changed, what conflicts, and what still needs a decision.
superseded
active_user counted on a 30-day window.
finance model · Q1
conflicts
Dashboard v1 still reports the 30-day count.
dashboard v1 · 3 places
✓ approved
active_user is the rolling 28-day definition for product reporting.
human decision · analyst · Jun 12
open
Which definition governs historical restatements?
awaiting review
See how Propperly works
The problem
Your agent can reach every source and still not know which answer is current.
Retrieval finds the documents. It doesn’t tell the next analyst — or the agent answering an executive — which definition is current, which decision replaced another, whether the numbers still conflict, or what was never resolved. That is not primarily a retrieval problem.
Current?
Which definition is the one to use right now.
Superseded?
What replaced what — and why it changed.
Conflict?
Whether sources still disagree with each other.
Unresolved?
What the team has never actually settled.
What Propperly does differently
Propperly maintains the conclusion, not just the documents.
Retrieval can hand your agent every file that mentions active_user. Propperly tracks how that one definition actually evolved — so the agent reads today’s answer, with the history behind it.
active_user · revenue analyticstracked over time
- superseded
A 30-day window was the working definition.
finance model · Q1 - challenged
A new analysis argued the 30-day window overstated active users.
analysis · May - current
The rolling 28-day definition was approved for product reporting.
human decision · analyst · Jun 12 - conflict
Three dashboards still report the old 30-day count.
dashboard v1 · 3 places - unresolved
Which definition governs historical restatements is still open.
awaiting review
The agent doesn’t just find every doc that mentions active_user — it knows which definition it’s allowed to rely on, and why.
Start small · Bounded Reconstruction
Test it on one analytics domain before you integrate.
You don’t need a data migration to see the value. Point Propperly at the work your team already has for a single domain, and it reconstructs what looks current, superseded, conflicting and unresolved — for your team to review.
- metric definitions
- dashboards
- analyses
- SQL & notebooks
- decisions
- docs & tickets
The Local Evaluation Kit runs on your side, on work you choose — offered as a hands-on evaluation, not a public download.
Engagement model
Start with the context. Keep it current.
Two stages of working with Propperly — first build the reviewed context your analytics agent relies on, then keep it current as the organization changes. Start on one bounded domain.
Managed Context Maintenance
How it works
Propperly maintains the state of the work.
01 · Reconstruct
Read what exists
Propperly reads the work already there — documents, analyses, code, AI sessions — and rebuilds the project's current understanding.
02 · Govern
Your team decides and attributes
Someone confirms, corrects, or rejects it. That decision is attributed and persists — downstream AI inherits governed state, not raw inference.
◆ human authority required03 · Use
Build on it
People and AI agents read the approved state directly: decisions, evidence, open questions, ruled-out paths, sources.
04 · Maintain
Keep it current
When sources change, recommission the domain. Propperly surfaces relevant differences and routes what requires renewed human governance — the governed state updates without rewriting its history.
26–27 points above two summary baselines · retrospective Mozilla evaluation (family-macro TAR)
Evidence, not another AI promise
Measured on a sealed corpus.
On a retrospective Mozilla evaluation over a sealed longitudinal corpus, Propperly's maintained state scored 26–27 points above two summary baselines on family-macro TAR. Those summary-arm differences were statistically robust in the retrospective evaluation, though the summary baselines had important representation limitations. Its 9.76-point advantage over a historical hybrid-RAG system was not statistically conclusive. Built by an analytics leader with a decade of experience, including leading an analytics team, and currently evaluated with analytics teams in B2B data and media.
Research
Evidence-led writing on maintained understanding.
How a project loses its own history, why AI memory is not a maintained state, and what analytics agents need beyond retrieval.
Product thinking
Browse the research library →The Picture Keeps Moving
Analytics practiceCan AI Replace Dashboards? What Changes—and What Doesn't
Analytics practiceWhy AI Analytics Still Gives Different Answers to the Same Business Question
◆ Also exploring · Labs
About the founding beta →Project Continuity — keep long-running Claude Code projects from losing their history.