AI Analyst· Conversational strategic analysis
Search & commands⌘K

What should we make of the position?

Ask in plain language. The analyst reasons over your signals, scenarios, and past decisions, and never publishes without your call.

Analyze a scenario
Pressure-test likelihood, desirability and achievability against live signals.
→
Map emerging signals
Cluster this week's 240 signals into the shifts that actually move a scenario.
→
Stress-test a strategy
Run the current plan against three adversarial futures and surface where it breaks.
→
Draft a board brief
One page: what changed, why it matters, and the decision it asks of you.
→
The analyst can be wrong. Every consequential change is reviewable, reversible, and stops for your approval.
Signals 3 agents working

The board's watching. The agents are working.

Agents triage signals around the clock and update scenarios within the limits you set. The consequential moves stop here for your call.

Signals triaged today
240
94% handled autonomously
Awaiting your review
03
1 endstate rewrite · oldest 12m
Scenarios updated
17
within governance policy
Analyst hours saved
31h
vs. manual synthesis baseline

Agent activity

live · last 30 min · 5 events
14:31
Analysis agent
Wants to rewrite the Green Acceleration endstate to Sustainable Market Leader after connecting 4 new events. Review changes →
match 92% · scenario green-acceleration
Awaiting you
14:22
Signal agent
Clustered 240 signals into 6 shifts; promoted battery passport rules to high-impact.
confidence 96% · policy: signal-triage = auto
Auto
14:09
Jane Doe
Approved analysis agent's downgrade of Fast Fossil Retreat likelihood to 41%.
agent-prepared · human-run
Approved
13:56
Analysis agent
Enriched scenario #GA-7 with 12 sources; achievability held at 72%.
confidence 89% · read-only enrichment
Auto
13:38
Jane Doe
Sent back an agent request to publish to The Wall. Wanted a second source on labor realignment.
fed back · won't re-propose without evidence
Sent back
Canvas / Green Acceleration agent proposed a change
12 / 12 events connected

Review agent changes

The analysis agent rewrote an endstate after connecting four new events. Approve it, revise it, or send it back before anything publishes.

Green Acceleration→Sustainable Market Leader
Sustainable Market Leader
Analysis agent rewrote this endstate after 4 new events connected
Likelihood
86%+8
Desirability
86%+1
Achievability
68%+6
Events feeding this endstate
2027
Battery passport rules
EU mandate makes bankable reuse contracts the default; favors domestic build policy.
2026
Circular feedstock auctions
Secondary-material auctions clear at scale; reinforces affordability crosswinds.
2026
AI infrastructure costs settle
Frees engineering spend that later firms redirect into platform work.
2025
Passport rules → domestic build
Policy, software and circular supply begin binding into one operating posture.
Analysis agent1 / 3
Does this still represent our strategic view?
The agent rewrote the endstate and 3 scores. Nothing publishes until you approve.
Reasoning92% confidence
  1. Causal chain tightened. 4 new events connect policy → domestic build → circular supply into one path, not three loose bets.
  2. Likelihood moved most. Battery-passport rules make the affordability crosswind bankable, lifting likelihood 78 → 86%.
  3. Narrative rewritten. Endstate now reads as a default operating posture by 2028, across Political, Economic and Technological dimensions.
What will change4 modified
−endstate = "Green Acceleration"
+endstate = "Sustainable Market Leader"
−likelihood = 78%
+likelihood = 86% · events 8 → 12
Two approvals required. Policy locks endstate rewrites to a human, and publishing to The Wall needs a second strategy-lead sign-off.
reversible · reverts to prior endstate if not published in 72h
Decisions / Sustainable Market Leader endstate 3 in the room

No one ships to The Wall alone.

Publishing a rewritten endstate takes two strategy leads. The team decides here, the analysis agent stays in the room to answer for its reasoning, and every word is on the record.

AA
Analysis agent automated14:31
I've rewritten the Green Acceleration endstate to Sustainable Market Leader. 4 new events tightened the causal chain and lifted likelihood 78 → 86%. Publishing to The Wall needs two strategy-lead approvals.
JD
Jane Doe strategy lead · admin14:33
The causal chain reads tighter than our last version: policy → domestic build → circular supply as one path. I'm approving. Looping in @Marcus for the second sign-off before it hits The Wall. approved
MO
Marcus Osei strategy lead14:35
Agreed the story is cleaner. @Analysis agent how sensitive is the 86% to the battery-passport rules holding? If the EU mandate slips a year, where does the endstate land?
AA
Analysis agent automated14:35
Battery-passport rules contribute +5 of the +8 pts. If the mandate slips a year, likelihood settles at ~81% and desirability is unchanged. It stays the leading endstate, though I'd re-open the review at that point. Full sensitivity in the canvas ↗
MO
Marcus Osei strategy lead14:36
That's well within tolerance. Approving — ship it to The Wall. approved
the agent answers in-thread
Monitoring· Decision ledger

Every agent move, on the record.

An append-only ledger of what the agents did, what they proposed, and every call a human made, with the policy that governed it.

All
Awaiting
Auto
Human decision
Sent back
Time
Actor
Action
Outcome
14:31:06
Analysis agent
Proposed endstate rewrite · Green Acceleration → Sustainable Market Leader
Awaiting
14:22:41
Signal agent
Clustered 240 signals · promoted battery-passport rules to high-impact
Auto
14:09:12
Jane Doe
Approved likelihood downgrade · Fast Fossil Retreat → 41%
Approved
13:56:03
Analysis agent
Enriched scenario #GA-7 with 12 sources · read-only
Auto
13:38:55
Jane Doe
Sent back publish request · required a second source on labor realignment
Sent back
13:20:17
Signal agent
Ingested 62 signals from 9 sources · deduped 14
Auto
Governance· Agent policy

You draw the line agents can't cross alone.

One place decides what runs autonomously and what stops for a human. This policy is what every "Auto" and "Awaiting you" tag on the other screens obeys.

Autonomy rules

4 active
Scenario score change under 10%
Small likelihood / desirability / achievability moves.
Auto-publish
Signal triage & clustering
Ingesting, deduping and grouping raw signals.
Auto-publish
Endstate rewrite
Changing what a scenario ultimately means.
1 reviewer
Publish to The Wall
Anything the board and org will see.
2 reviewers

Confidence floor

Agents may act autonomously above 90% confidence
Below the floor, the agent must route to a human even for auto-class actions.
Floor: 90%