Surface 04 · OperationsReplaces the job of: n8n · Zapier · Linear · Asana · Notion · ClickUp · Looker · GA4 · Northbeam · Triple Whale · Fivetran · SegmentLive in all deployments

See what is waiting, running, blocked, approved, and done.

Operations shows the live work: automations, approvals, owners, dashboards, goals, and exceptions. When a job fails or a decision is waiting, the team sees what happened and who needs to act.

Read the docs
What this view gives you

Live workflows + approvals

Every automation is visible, every job has a status, and approvals happen before risky changes go live.

Goals tied to work

Goals, owners, and metrics stay linked to the campaigns, tests, workflows, and tasks that move them.

Charts tied to the live work

Charts and reports stay connected to the launch, customer, order, approval, or workflow they explain.

Attribution that holds up

Channels and tests are tied back to orders, spend, and revenue where the data is available.

Decisions with context

Every decision shows the work behind it: the numbers, source files, approvals, and next step.

First-party data when needed

Tracking, connections, reconciliation, and visualization can run inside the same workflow instead of becoming a separate data project.

Work this can reduce
TaskSimilar toolsLyberty
Workflow automationn8n / Zapier / Retool ($1-3k/mo)included
Project & OKRsAsana + Linear + Notion ($30-60/seat)included
Dashboards & BILooker + GA4 ($3-8k/mo)included
Attribution / analyticsNorthbeam + Triple Whale ($1-3k/mo)included
Data pipelinesFivetran + Segment + GTM ($2-6k/mo)first-party capable
Past decisionsNotion + Glean ($30-50/seat)included
01 · See every run in one place

A live work queue for what is running, blocked, and waiting.

Orders · approvals · jobs · automations

In flight
4
Awaiting approval
2
Failed · 24h
1
Scheduled
12
Median run
184ms
operations.runs — live · 8 shown · filtered: active + 24h⌘K to command
WorkflowStatusDurationStepActor
fulfill_order#wf_0412
● running
00:00:04
3/7 · notify_carrier
workflow · fulfillment
reconcile_payouts#wf_0398
● awaiting approval
pending 14m
approval · nia@
human · finance
import.hubspot.contacts#wf_0391
● succeeded
02:14.112
7/7 done
workflow · sync
audience.recompute#wf_0388
● running
00:12.004
2/4 · match_customers
workflow · audience
ledger.close.monthly#wf_0380
● scheduled
in 2h
awaiting trigger
cron · 0 18 * * *
test.promote_winner#wf_0376
● needs review
42m
4/5 · decide_next_action
human · david@
stripe.charge.backfill#wf_0362
● failed
00:00:18
2/3 · approval_missing
workflow · finance
launch_runbook_q2#wf_0341
● succeeded
01:02.482
12/12 done
workflow · release
02 · The difference

Zapier · Retool · Workato gives you the tool. Lyberty ties it to the rest of the business.

Under the hood

Zapier · Retool · Workato
Automations that run somewhere and leave logs you hope are enough.
TriggerWebhooks and scheduled jobs against each tool
Run logPer-tool logs, kept 30-90 days
RulesRules live in configs per tool
ApprovalsSlack message, screenshot, spreadsheet
DebuggingOpen 4 tools, compare timestamps
Lyberty
Automations that leave evidence, stay linked to the order, and show who approved the risky steps.
TriggerOne work feed, one place to look
Run logEach run keeps the steps and evidence you need
RulesApprovals and guardrails checked before action
ApprovalsApprovals linked to the work
DebuggingClick the run and see what happened
03 · Automations with business context

Every automation can see the order, customer, approval, and rule it touches.

Not scripts in a vacuum

Workflow: reconcile_payouts · 5 steps · DAG
stripe.payoutpolicy.gatematch.orderpost.entrysave.evidence
External fact
Policy gate
Action
Evidence saved

Each step checks the right guardrails. Each result is recorded against the right order, customer, campaign, or finance line. When a run ends, you get evidence you can inspect and share.

This is not a nicer Zapier. It is a way to run the workflow and keep the approval, result, and evidence attached.

04 · Commands you'll actually use

One keyboard shortcut for the work people keep asking about.

⌘K

▸ run fulfill_order(order_id=ord_9f22)// kick off a workflow run
▸ approve wf_0398 --note "looks good"// clear an approval, signed as you
▸ retry wf_0362 --from step=2// resume from a specific step
▸ save reconcile.monthly --period=2026-03// save close evidence
▸ inspect party:acc_meridian --trail// see everything that happened to an object
01 · Plans tied to real work

The plan should show what shipped, what slipped, and why.

Not a deck · not a spreadsheet

plan.objective — obj_q2_nrr55% on track · 3 commitments
objGrow NRR from 102 → 110
target: +8pp
commitEnterprise QBR program
8 accounts · Q2
wfworkflow: qbr.cycle
6 runs · 2 awaiting
measuremeasure.nrr_enterprise
107 → 109.4
commitPricing V3 rollout
Jun 30
artifactcanvas: pricing v3
rev 12 · linked 4
expexperiment.pricing_v3
82% conf · holding
commitOnboarding rework
Q2 · engineering

Every objective links to the work that moves it, the tests that challenged it, and the numbers that prove it. "How are we doing on Q2?" should not require a screenshot update.

When something slips, you see it the day it slipped — not in the QBR three weeks later.

02 · The difference

Notion · Asana · Linear roadmaps gives you the tool. Lyberty ties it to the rest of the business.

Under the hood

Notion · Asana · Linear roadmaps
A tidy list of goals that can drift away from what the team is actually doing.
ObjectiveA doc, a row, or a label
ProgressA slider a human drags weekly
CommitmentsA list, maybe synced
ReviewMeeting + screenshots
DriftQuiet, over months
Lyberty
A plan tied to the orders, tests, workflows, and decisions that move it.
ObjectiveLinked to the work and the numbers behind it
ProgressCalculated from real work where possible
CommitmentsTracked with owner, status, and proof
ReviewOne link to the work and evidence
DriftVisible the day it happens
01 · Decisions with evidence

When Lyberty suggests a next move, you can see why.

Not another chat box

decision history — Q2 · 42 entries6 decisions · 3 playbooks · 8 hypotheses
Decision: shut paid social
2026-04-08

Blended CAC ran 4.2x LTV over 30 days. Cohort decay confirmed in the revenue readout. Suggested by the workflow, approved by Nia.

financegtmevidence: ledger result · experiment 6
Playbook: enterprise QBR cycle
2026-04-02

Three wins in sequence used the same follow-up pattern. The team saved it as a reusable playbook.

gtmpatternevidence: workflow 14 · outcome 104
Decision: hold pricing V3
2026-03-27

Churn signal appeared in cohort c_3. Decision: hold until July and collect more data.

pricingriskevidence: hypothesis 22 · cohort notes
Insight: onboarding dropoff at step 4
2026-03-15

Experiment B confirmed the issue. Step 4 was rewritten. Dropoff fell 38%.

productevidence: experiment 4 · cohort 12

Every decision leaves a trail. Every hypothesis keeps its evidence. Every suggestion points back to the test and the number behind it. The next person can see what the team already tried.

What matters is simple: when the same question comes up again, the team can see the last answer, the proof, and the next action.

02 · The difference

Glean · ChatGPT Enterprise · your wiki gives you the tool. Lyberty ties it to the rest of the business.

Under the hood

Glean · ChatGPT Enterprise · your wiki
A search bar over documents the company happened to write down.
AnswersFrom docs you hope are current
RecommendationsConfidence score with no sources
DecisionsLive in Slack threads
LearningYou hope people read the wiki
ReviewReconstruct from emails
Lyberty
Answers tied to the actual work: tests, orders, decisions, and changes.
AnswersFrom current work and source evidence
RecommendationsLinked to evidence you can verify
DecisionsFiled with the reason behind them
LearningWinners can become reusable playbooks
ReviewEvery answer shows its work

The surface is just the view.

All six surfaces share the same record. Adopt one, and the others are already reading from what you wrote — no sync, no export, no reconcile.

See every surface