RnBy v2 · Strategy review

RnBy

One company. Single prompt.

The control plane for your AI stack. Agents do the work. You approve the outcomes.

Apps & Tools Data Silos Foundational Models R RnBy
Ask
Type, talk, or screenshot.
Route
The right agent picks up.
Prove
Every step is logged.

01 · The Buyer Problem

Owners want execution, not more AI tools.

Stop managing tasks. Start owning outcomes with a deterministic, private autopilot.

01

Setup gets stuck

Accounts, forms, and follow-up spread everywhere.

02

Tools hurt trust

Each fix asks for another login or prompt.

03

No supervision

People need to see what changed and what is next.

Directional signal

More setup means less operator confidence.

Simple start More tools Setup fatigue High Low Operator Confidence Setup Burden

RnBy wins by absorbing the mess.

Plain language Hidden complexity
02 · The wedge

Start with the three things you do every day.

Inbox, sales follow-up, and back-office ops. Daily, visible, universal — automate these first. HR and IT come when trust is earned.

90%
Urgency
86%
Clarity
82%
Proof
88%
Expansion
The wedge · three daily domains

From scattered replies to finished work.

BEFORE: DAILY GRIND Emails pile up Leads go cold Invoices slip Meetings miss OT AFTER: AUTOPILOT Day cleared Auto-run by noon

03 · How it feels

Intent transformation.

Ask naturally

Type, talk, or send a screenshot.

Route backstage

The right model handles the messy part.

Show proof

Tasks, owners, and status stay visible.

Move on fast

The owner approves, learns, or redirects.

Owner effort -42% less setup drag
Visibility 4 / 4 status checkpoints
Trust Live approvals and proof
Screenshot-first help Visible next step No AI stack knowledge required

04 · Expert Agents

The Swarm: Specialized digital employees.

Deploy autonomous agents for HR, IT, Sales, and Finance under one CEO Orchestrator.

HR & Talent
People, apps, and day-one readiness.
IT & Systems
Processes, proof, and exception handling.
Sales & CRM
Web, content, and customer-facing work.

The swarm operates as a cohesive unit. Specialized agents continuously talk to each other to share context, while the CEO Orchestrator correlates information to drive final execution.

Memory Core IT & Systems Sales & CRM Legal & Finance HR & Talent R CEO Orchestrator
05 · Product Rules

Four pillars of Neural Infrastructure.

Deterministic

Consistent and reliable execution without hallucination. Non-random output.

Private

Your proprietary data stays yours. SOC2 compliant and never used for global training.

Autonomous

Agents operate on autopilot in the background, executing complex chains of tasks.

Audited

Full transparency and governance. Every intent, step, and decision is logged.

Deterministic Private Autonomous Audited

Chapter 06 · Revenue

Recurring revenue overtakes setup fast.

Setup creates fast cash. Managed ops compounds it.

Advisory fee Implementation fee Managed subscription Referral / reseller upside
Setup & Advisory Managed Ops Total Revenue $20k $14k $2k $4k $60k $24k $62k Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 15× managed ops in 6 months
06b · Packaging logic

Structured for rapid expansion.

Consumer

Fast first sale. Frictionless entry.

CORE TIER

Business

Core subscription and managed operations.

Expansion

Higher-value automation and specialized agents.

STICKY REVENUE Setup Monthly Ops Premium

Chapter 07 · Go to market

One platform, two entry points.

Sell through founder pain now. Expand through proof later.

1
Design partners

3 to 5 owners with visible onboarding pain.

2
Founder-led win

Do the first workflow with them.

3
Template it

Turn wins into repeatable launch paths.

4
Expand carefully

Open adjacent workflows after trust.

Proof Expansion (Time) Founder Pain Founder Pain Initial Proof Scale / Expansion

08 · The Moat

The Memory Core: A private knowledge graph.

Your data stays yours. SOC2 compliance with dedicated vector namespaces.

MemoryMore company knowledge is ingested.
Role expertiseEach agent learns its company-specific job.
GuidanceRecommendations get sharper and faster.
FutureGrounded agent meetings become useful.
Memory Core Richer Memory Sharper Agents Better Guidance Agent Council

Chapter 09 · Roadmap

30 days from alignment to first live win.

A rolling 4-week plan distilled from the Shaun session. Each week has an outcome gate — we don't advance until it clears.

Week 1

Lock the v2 thesis.

  • One-sentence v2 thesis signed off by Matt + Shaun.
  • 30-day primary customer chosen: enterprise onboarding vs consumer pilot.
  • v1.5 → v2.0 scope boundary published: in, required, deferred.
  • Shaun session ingested and retrievable by role agents.
Outcome gate Written thesis + target customer + scope doc.
Week 2

Non-technical onboarding.

  • Happy-path flow for a 65-year-old CEO — no API-key blockers.
  • BYOK vs managed-keys vs guided-checklist decision recorded.
  • First task-to-model routing matrix (websites, images, support, analysis).
  • Screenshot-first help flow: "I'm stuck" → upload → guided recovery.
Outcome gate A non-technical user reaches first value without help.
Week 3

Package the offers.

  • Consumer offer live: $100–$500 setup + light coaching.
  • Enterprise offer: onboarding + role agents + managed AI stack.
  • Pricing validation with 3–5 early adopters, real commitments.
  • Mobile strategy chosen: Telegram, lightweight web, or native wrapper.
Outcome gate At least one paid consumer setup booked.
Week 4

Ground wave — first wins.

  • 3–5 design partners running live workflows end-to-end.
  • Ingestion expanded: emails, ops docs, internal notes.
  • Role-agent grounding requirements documented per role.
  • Swarm-meeting prototype demoed with grounded company context.
Outcome gate First customer says "keep it running."
By day 30

Proof, not promise.

3–5 design partners live
$1k+/wk consumer revenue hypothesis
1 published v2 scope doc
100% ingestion of session memory
10 · Decision today

Decide three things now.

Primary buyer

Founder-led SMBs with obvious workflow pain.

Product thesis

AI operations layer first.

Build focus

Usability, proof, and repeatable ROI.

Success looks like
A non-technical owner can hand RnBy a messy workflow and say, “Make this easy.”
Simple to buy Simple to trust Hard to replace

Chapter 12 · Ecosystem

What's live. What's next.

The repo is past placeholder integrations: tenant sync, managed email, AI routing, and operator controls are already real. The next work is reliability hardening plus earned external surfaces.

Live now

Implemented · tenant + control planes

AI + platform
Claude OpenAI Gemini Tenant DB tools
Knowledge + work
Google Drive GitHub Slack ClickUp Spinach
Comms + platform
Google Workspace Personal Google Resend Supabase

Hardening now

Runtime maturity · proof path

Operator reliability
Fleet deploy parity Migration checks Task recovery Queue visibility

The product proof is reliability: agents can work across tenants without losing state, approval context, or deployment parity.

Next horizon

Earned after trust · not yet core

External work surfaces
Outlook / Teams Payments
Product surfaces
Portal runtime Website automation CRM workflow depth
Agent frontier
Evaluation layer

Chapter 13 · App concepts

One company. One screen.

Owner cockpit Use the real RnBy pattern: dark rail, command header, task board, and visible approvals.
Task board

Tasks

Create work, assign an owner, and keep progress visible.

42Open tasks 7Waiting 5Live workers

Needs review

Approve onboarding email Inbox agent · customer-ready draft

In progress

Google Drive sync Chunk 4 running · heartbeat live

Recovered

Gmail cleanup completed Earlier failed duplicate superseded
R RnBy Cockpit Agents Integrations Help & Support YOUR DAY · FRI 24 APR 3 things need your eyes today. SWARM RUNNING 01 · AGENTS Three operators on shift INBOX AGENT Drafting replies LIVE 24 drafts ready · 3 await your approval Gmail Slack Outlook SALES AGENT Working pipeline LIVE 12 follow-ups sent · 4 leads warm today HubSpot Sheets Stripe OPS AGENT Filing & scheduling LIVE 8 invoices queued · all scheduled Calendar Drive GitHub 02 · TASKS Live work · queued, running, done NEEDS REVIEW · 3 IN MOTION · 5 DONE · 16 Reply to Sarah on pricing Inbox agent · drafted 2m ago APPROVE Quote draft for Acme #2218 Sales agent · drafted 18m ago APPROVE Onboarding email · Innovate Co Inbox agent · sending now Follow-up batch · 4 cold leads Sales agent · sequencing Acme quote #2218 — sent Sales agent · 2h ago Invoice #501 Greenleaf — scheduled Ops agent · runs Mon 7am Vendor reply · Logistics Co. Inbox agent · 5m ago 03 · PROGRESS Velocity climbing HOURS SAVED · 7 DAYS 18.4 ▲ 41% vs last week PEAK Mon Tue Wed Thu Fri Sat Sun TRUST CLIMBING First-time success 91% Auto vs escalated 7 : 1 Knowledge freshness 2.1 d
Task flow
Ask, approve, move on.
Needs review In motion Done

The box should never trap the owner. Questions stay visible, approvals stay obvious.

Expert agents
Three agents, one control plane.
Inbox Sales Ops

HR, IT, and coaching join once trust is earned on the daily three.

14 · Agent hierarchy

One CEO. Three operators. One swarm.

Intent flows down to the agents. Memory and outcomes climb back up. The CEO never has to do the work — and the operator never babysits the tools.

CEO ORCHESTRATOR Sets intent. Holds standards · routes work. INBOX AGENT Reads, drafts, replies. Customer email · vendor questions · lead replies. REACHES Gmail Slack Outlook SALES AGENT Routes, follows up. Leads · quotes · pipeline · CRM · meeting prep. REACHES HubSpot Sheets Stripe OPS AGENT Schedules, files. Invoices · calendar · docs · books · payables. REACHES Calendar Drive GitHub SHARED MEMORY · GOOGLE WORKSPACE · CLAUDE · OPENAI · GEMINI Every agent borrows from one knowledge core.
15 · Operator telemetry

Autopilot, not auto-theatre.

Four numbers tell the owner whether trust is climbing.

First-time success 91% Zero operator edits.
Time-to-first-value 38 min New tenant to first win.
Knowledge freshness 2.1 d Since last ingestion.
Auto / escalate 7 : 1 Closed without humans.
R RnBy OPERATOR TELEMETRY 91% First-time success 38m Time to first value 7 : 1 Auto / escalated Tasks auto-resolved / hour -24h Now Approvals in queue 3 low 12 now Knowledge coverage — tenant & agent-specific (30 days) Day 0 Day 30 Tenant docs Role memory Today's queue Inbox agent 24 drafted · 3 need you REVIEW Sales agent 12 follow-ups · 4 warm leads RUNNING Ops agent 8 invoices · all scheduled CLEAR Trust climbing — 30 days Day 0 Day 30 7 : 1 auto 51% first-time → 91% first-time
Auto-resolved Awaiting operator Knowledge coverage
v2 rollout

Capability rollout

Shipped · in motion · backlog. Moves forward on proof, not promise.

Onboarding wedge
66% live
Expert agents
56% live
Knowledge freshness
56% live
Screenshot help
40% live
Voice + swarm meetings
24% live
Why this matters

Every number above equals a reason owners keep paying.

One more thing

The deck ends here. The product starts Monday.

3of 5 design-partner slots open · v2 cohort

RnBy v2 · 2026 matt@rnby.ai