Data sources the assistant reads from. Disconnecting removes a source from context immediately; ingested data is retained.
SOURCESsorted by status
AI Memory
Persistent context injected into every prompt. Treat these as a commonplace book — short, opinionated, kept current. Files are concatenated in the order shown.
Files
—
Total tokens
—
Budget
—
Last refreshed
—
File name (lowercase, no spaces)
Description (one line)
Content
Add task
Task
Project
Kind
Action
€165k
Current baseline
€500k
18-mo target ARR
€570k
Base case exit
10
FMO signings needed
6.3×
LTV : CAC
4.8mo
CAC payback
Key Actions
Your weekly unblocking list. Check off what's done — state saved locally. P1 = this week · P2 = this month · P3 = this quarter.
The core bet
The US life/health/annuity distribution market is large, fragmented, and tech-hungry. Roughly 500–1,000 FMOs and IMOs have the scale and budget to adopt operational software. Cross-sell remains the most under-executed revenue lever in the channel — and it's the lever whose economics align perfectly with performance-based SaaS.
We are not a SaaS company chasing volume. We are a performance-revenue business with a SaaS front door. 10 deep, referring FMOs beats 100 shallow, price-sensitive ones. Every decision — pricing, targeting, messaging, delivery — filters through that lens.
Three-product stack
RevEngine SaaS
Wedge
Tiered $295–$4,995/mo based on FMO size. Policy analysis, cross-sell opportunity identification, agent prioritization. 30-day trial. Annual billing saves 15%.
RevEngine Campaigns
60–70% of revenue
$495/agent/campaign + 25% Y1 commission + 10% renewals Y2–3. Direct response gap campaigns standard; referral generation & nurture as upsells.
Kevlex Digital retainers
Passive baseline
€12.5k/mo existing retainers delivered by Mitch. Not actively sold. Reactivated only as white-glove upsell ($7,500/mo floor) to RevEngine enterprise clients.
MyInsuranceCoach
Shelved
Website footer only. No marketing spend. Kill if no meaningful inbound in 6 months.
What wins this plan
Wedge-and-expand — SaaS trial gets the CEO to yes on the easy thing first; Campaigns is the follow-on conversation 30 days later.
Performance alignment — commission share means FMOs only pay meaningfully when they make money. CEOs have personal override income in the loop — they self-police attribution.
Platform-agnostic delivery — n8n backend stays proprietary; connects to GHL, AgencyBloc, Radius. Kevlex keeps the IP; FMOs keep their CRM of choice.
Referral density over volume — aim for 8–12 anchor FMOs, each generating 2–3 warm referrals, not a thousand-FMO long tail.
Quarterly "State of FMO Cross-Sell" benchmark report
Biweekly newsletter, 800–1,200 words
LinkedIn daily presence
Events
~10% of pipeline
Build warm pipeline in volume.
One trade show month 6–8: Medicarians or ITC Vegas
Pre-event outbound to attendee list
20+ scheduled meetings on-site
Funnel math — what it takes to sign 10 FMOs
Stage
Conversion
Total (18mo)
Per month
Cold outbound contacts
—
~4,240
~280
Positive responses
5%
~212
~14
Booked meetings
25%
~53
~3.5
Pilot agreements
30%
~17
~1.1
Paid rollouts signed
60%
10
~0.7
Reality check: 280 contacts/month across ~500 named accounts = ~5–7 touches per account over 15 months. Kevin shows up for meetings and closes. An SDR or VA builds the lists and runs sequences starting ASAP.
Demo with CEO + ops lead (60 min) — live book analysis (anonymized KRS data), frame the pilot. Do not discount during this call.
Pilot proposal — 3 months, 1–2 agents, 3 Direct Response campaigns = $2,970 + SaaS at $295/mo discounted 50%. Lighthouse exception: first 2 clients free in exchange for case study rights.
Pilot kickoff — Kevin attends; Mitch and team deliver. CEO commitment addendum: weekly 30-min check-in during pilot.
Pilot review day 75 — present lift metrics, pitch rollout, ask for 12-month annual commitment with full agent complement.
Partnerships
Matt Ruddick / KRS
Matt paid to build most of what became RevEngine. KRS holds all downstream appointments — single contract, single fee flow.
Kevin's freedom
Unrestricted — RevEngine is Kevlex IP
KRS exclusivity
Flagship Medicare case study; no competing AZ FMO 50+ agents without notice
Referral fee
10% of Y1 revenue on warm intros from Matt
Advisor role
"RevEngine Advisor" title + quarterly check-in
Action
One-pager this week. Formalize before FMO #2.
Randy Schwantz / The Wedge Group
Commercial P&C agencies, sales coaching focus. No direct overlap. Partnership unlocks commercial TAM without Kevin building that motion from scratch.
Revenue split
60% Randy / 40% Kevin (his book is the asset)
Delivery split
Kevin technical; Randy sales + coaching
Min commitment
2 paying clients within 6 months, or partnership pauses
Gate
Formal only after month 6 of base-case personal lines progress
Growth Funnel
Full acquisition-to-revenue flow from outreach sources through steady-state FMO value. Volume and conversion KPIs at every stage.
Kevin's time allocation
The goal: shift from 70% executing to 60% selling by Q4 2027. Every brief written, playbook handed off, and hire made buys back strategic hours.
Kevin now
Executing70%
Designing20%
Selling10%
Kevin Q4 '26
Executing30%
Designing30%
Selling40%
Kevin Q4 '27
Executing15%
Designing25%
Selling60%
Delegation model
Mitch
Operations Lead → Delivery Lead
Owns
Campaign execution + client Slack/email
Hatch management + agent support
Delivery playbook maintenance
Client onboarding (post-handoff from Kevin)
Kevin's rule: If Mitch needs Kevin more than 3× per week on the same topic, write the SOP instead of answering again.
Slow factor: Give Mitch 5× Kevin's estimated time on any new task.
Joe
Technical Lead
Owns
GHL integration maintenance
Portal technical support
Automation building (from Kevin scaffolds)
Railway deployments + RevEngine MCP ops
Kevin's rule: Kevin builds version 1. Joe extends and maintains. Always. No exceptions.
Slow factor: Joe needs the full spec, working code sample, and Railway credentials before starting.
Kevin's rule: Give Eillyne a template for everything. Creating the template is Kevin's job.
Slow factor: 8× time buffer. Invest now — pays back from month 6 onward.
Hiring plan
Role
When
Budget
Gate
SDR / VA
Month 2–4
€1k–2k/mo, 20–30 hrs/wk
Start asap — protects Kevin's sales time. Remote EU or LatAm.
Customer Success lead
Month 5–7
€4k–6k/mo
Conditional: go/no-go end of month 4 based on pilot-to-paid rate.
2nd SDR
Month 14+
TBD
Evaluate performance of first SDR at month 13.
The brief-writing habit. Before every task you're about to do yourself — ask: "Could Mitch or Joe do this in a week if I wrote a 10-minute brief?" If yes, write the brief. The brief takes 10 minutes. You save 2–4 hours. Over a month, that's the difference between 40 and 100 Kevin-hours of strategic capacity.
Kevin's weekly schedule
30 hours/week. 50/50 split acquisition/delivery, adjusting to 70/30 toward acquisition as Mitch takes over delivery (months 3+).
Updates project phase & summary from fresh touchpoints. Generates deliverables for projects that have none.
Inbox Tools
Scans all open rows for missing fields (title, summary, kind) and fills them in using Haiku. Fast and non-destructive — nothing gets deleted or rewritten.
Checks every open task against the last 24h of email, Slack, and meetings. Updates context notes and flags items that look resolved — but never changes Status; you confirm those.
Full inbox sync: runs the morning brief to find new items across all channels, then immediately runs the executor to process the entire queue. Slow — one Sonnet call per row.
Plan & Blockers
Re-runs the weekly plan refresh using the last 48h of signals. Updates section summaries, touchpoints, and drift notes. Rate-limited to once per 10 minutes.
Scans the last 24h of signals for new blockers not yet in the list, and updates descriptions on existing ones if the signal has changed. Uses Haiku.
Other Tools
Runs the morning brief agent only — scans Gmail, Calendar, Slack, and Read.ai, then creates new Processing rows. Does not run the executor; rows stay in Processing until you do.
Runs the executor agent only — takes every existing Processing row and generates its output (email draft, calendar event, etc.), moving each row to Ready. Does not find new items.
Sends a test push notification to your phone via ntfy.sh. Use this to verify your notification topic is configured correctly and working.
Reads the last 24h of touchpoints for every active project and updates each project's summary, phase, and last_signal_at. Runs automatically on weekdays at 10:30.
Generates a weekly retrospective from completed and dismissed items of the past week. Normally runs automatically on Friday at 17:00; trigger manually if you want it mid-week.
Scans open Ready items against last 24h signals and flags each as Probably-Done, Probably-Dismiss, or Probably-Stale. Runs automatically at 10:00 and 16:00 on weekdays.
Generates AI-suggested deliverables for active projects that don't have any yet. Runs automatically on Sunday at 21:55. Trigger manually after adding a new project.
Runs meeting prep for every event on today's calendar. Each one writes a Ready inbox row with prep notes and sends an ntfy push. Runs automatically at 08:00 (SOD) and 13:00 (MOD).
System prompt
Loading…
User Management
Users who can log in to this EA instance.
Email
Display name
User
Role
Timezone
Setup
Loading…
Configure · The Ledger
LLM Cost — accounts for the press
What the agent spent and where. Caching reduces input tokens; reasoning-heavy work goes to Sonnet, lighter classification to Haiku.
Range
Users
01
By function — the agent's call sheet
CallerCallsInput tokOutput tokCacheCostShare
02
By model
ModelCallsCostShare
03
Daily spend
DateCallsCostSpend
End of ledger · figures provisional until end-of-day reconciliation
Developer
API Health
Loading…
Environment
Loading…
Action Counts
Loading…
Integrations
Loading…
Changelog
Version History
Loading…
Daily Feedback
Loading…
—
Follow-upsEdit before confirming
Add Blocker
Title
Description (optional)
Priority
Project
⚠️API tokens exhausted
The morning brief and scheduled tasks couldn't run because the Anthropic API ran out of credits.
Top up your balance, then click Initialize to run the brief now.
Error details
New project detected
The agent filed inbox items under a project name that doesn't exist yet:
Add as new project
Project name
Context (1–2 sentences for the agent)
Assign to existing project
Renames the project field on all matching inbox items.