Define the population
Make exclusions and assumptions part of context.
Them AI for Data and analytics
Agree on the meaning before debating the number
A metric can hide different definitions, exclusions, and decisions. Bring reviewed analytical context into shared rooms, so the team can inspect what a number means before building another report or asking an agent for analysis.
Mac-only for now. Free during the beta (invite-only).
Analytics roomShared room
You with your team
Why do these reports use different customer counts?
Them AI with company contextOne report counts active workspaces; the other counts paying accounts. The metric definitions describe different populations. Agree which question you need to answer before trying to reconcile the totals.
↗ Example metric definitionReview with Analytics lead
After human approvalCustomer population definition
A useful record for the next person.
Illustrative workflow. Not a live conversation.
The work gets better when the context stays.
Good judgment belongs
with the whole team.
Explain the difference
Compare definitions before trying to reconcile results.
Reuse sound questions
Keep analytical checks available to other teammates.
In the day-to-day
Work your team can bring to Them AI.
Start with the question. Keep the source. Make a person responsible for what comes next.
Data and analytics room Example 01
What exactly does active mean here?
Active means the account had a qualifying event within the reporting window, according to the glossary. Registered accounts are a different population and should not be substituted in the comparison.
↗ Example metric glossaryMetric definitions
Keep a metric's meaning beside the question it answers. Ask about populations, exclusions, and ownership, then cite the approved definition when a teammate needs to explain a report or comparison.
See how the workflow fits
- Bring
- Metric glossary
- Review
- Metric owner
- Keep
- Definition summary
Data and analytics room Example 02
What should we check before this analysis?
Confirm the population, comparison window, and handling of missing observations. The analysis method requires those choices to be explicit before an analyst runs the query and interprets the result.
↗ Example analysis methodVisible assumptions
Make analytical assumptions available before another person or agent repeats the work. Review the population, missing data, and interpretation limits in a room, keeping the actual query execution outside Them AI.
See how the workflow fits
- Bring
- Analysis methodology
- Review
- Analyst
- Keep
- Assumption checklist
Data and analytics room Example 03
Are these two reports measuring the same thing?
The definitions use different populations and reporting windows. Compare those choices first, then ask the reporting owners to validate the remaining difference in the underlying analysis.
↗ Example reporting notesReporting disagreements
When reports disagree, start with the reasoning behind each measure. Bring approved definitions into the discussion, identify what differs, and ask the relevant owners to resolve the interpretation before reuse.
See how the workflow fits
- Bring
- Definition notes
- Review
- Reporting owners
- Keep
- Difference outline
Data and analytics room Example 04
What should another analyst know before reuse?
Carry over the intended question, population, and known limitations. The handoff method asks the receiving analyst to confirm those assumptions before reusing the approach for a different decision.
↗ Example handoff methodAnalytical handoffs
Pass along the reasoning that a chart cannot show. Preserve the approved method, caveats, and unanswered questions, then make the review approach a skill another analyst or external agent can retrieve.
See how the workflow fits
- Bring
- Method and caveats
- Review
- Analysis owner
- Keep
- Handoff skill
These are example workflows to evaluate, not customer results. Connected agents use their own tools; Them AI is the shared context and review workspace.
From useful work to reusable knowledge
The conversation is
only the beginning.
Memory keeps the facts and decisions. Skills keep a way of working. People decide what deserves to become a shared reference.
Meet Rooms and the Collective- 01
Bring the context
Example metric definition
- 02
Make the human call
Analytics lead reviews the sources and proposed record.
- 03
Keep the method
Metric review
Make it real
Start with a disputed definition
Bring the approved glossary entry and the questions people ask when reports disagree.
Discuss a design partnership
Before you begin
A few things
worth knowing.
Does Them AI query our warehouse?
No native warehouse execution is promised. External agents can retrieve context through MCP and run their own analysis.
Should we upload raw customer data?
No. Human access is organization-wide today. Start with approved definitions and sanitized methods, not restricted datasets.
How can analytics try it?
Mac-only for now. Free during the beta (invite-only). Choose a shared metric and its responsible owner.
What does the current beta include?
Shared rooms, memory with sources and review status, reusable skills, and MCP access for connected agents. Mac-only for now. Free during the beta (invite-only). Human room and memory access is organization-wide. Client-scoped MCP feeds restrict agent retrieval to that client's approved records plus approved firm-wide material. Do not treat a room as a private boundary for individual members. Read the access boundaries.
Bring one real workflow
Make the next project
better informed than the last.
Start with a question your team keeps answering. Bring the sources, the people who know, and the decisions worth keeping.
Mac-only for now. Free during the beta (invite-only).