Research and outreach agents
- Prospect research filed before the workday starts
- Outreach drafted for human approval, never auto-sent
- CRM kept clean without anyone touching it
WAVERLY LAB · EST. 2019 · REBOOTED FOR THE AGENT ERA
Your business, running on agents.
Waverly Lab installs AI operating systems in founder-led businesses. No six-month engagement. No new software to migrate to. No slide deck.
Founded in 2019 as a product studio. Rebuilt for the agent era by Edgar Li after running a 20+ person operation serving 600+ clients.
THE BOTTLENECK
The founder should not be the operating system.
Founder-led businesses rarely stall because the team cannot do the work. They stall because every handoff, exception, and approval still routes through the founder. I rebuild that operating layer so routine work moves without you, while the decisions that need you arrive with context. You get leverage without losing control.
HOW IT WORKS
- 01
AI Operating Audit
One week. I map one workflow end to end (people, tools, handoffs, exceptions, and approvals), rank the automation opportunities with the ROI math shown, and put one proof live in your stack by Friday. You leave with the map, the demo, and an implementation plan your team can own. Under 3 hours of your time. The fee credits toward a sprint.
Book the audit$3,500
- 02
Implementation Sprint
Three to four weeks, fixed scope. I build the top systems from your audit and leave them running, with your team trained and every agent decision logged.
$12,500
- 03
Operating Partner
Agents run your ops layer. I run the agents. Continuous extension into new workflows, with oversight you can read. Pause anytime, unused time rolls over.
$15,000/mo
SYSTEMS I RUN
These run in my own company today.
I do not sell anything I do not run. Every system Waverly Lab installs is one already working in my own business, every day, in production.
Reporting agents
- Daily operating digest from live systems
- Numbers pulled, checked, and delivered before standup
- Exceptions flagged, not buried
A content engine
- Mines my knowledge base nightly and drafts in my voice
- Trend monitoring with takes ready for review each morning
- I approve every word that ships
SELECTED PROJECTS
Agent systems, operating platforms, and the businesses that taught me where human-only operations break.
Projects · Media
Presscart
Scaling a media business past 600 clients
B2B media and publications · Operations
View the project
Projects · Client work
Source Forward
A sales hub that trains reps and answers like the top rep
AI sales enablement · Energy services
View the project
Projects · Current
Sandbox
The platform where our agents live
GTM agent platform · Agent operations
View the project
Projects · Client work at Sandbox
UMS
An outbound pipeline that drafts, waits for a human, then sends
Agent operations · LinkedIn and email outbound
View the projectSee the complete track record across current operating work and the original product studio.
View all projectsTheir responsiveness and attention to detail is unlike anyone we have worked with before, and their process from ideation to deployment is something that helps us succeed day after day.
Owen Meyer · CEO, Liquor Lab
Founded in 2019 as a product studio, Waverly Lab has always worked the same way: map the work, test the smallest useful system, and leave the client with something their team can run. The tools changed. The method did not.
FAQ
What happens in the audit week?
Day 1 is a 90-minute deep-dive call. Days 2 through 4, I do the work: workflow analysis, ROI scoring, build planning, and a live proof-of-concept, with async updates instead of meetings. Day 5 is the delivery call: the full plan, the math, the demo, and exactly what I would do next in your shoes.
How much of my time does this take?
Under 3 hours for the audit week. Sprints need a similar amount per week, mostly in approvals. Your team keeps doing their jobs the whole time.
Do we need to change our tools?
No. Agents are built on top of the stack you already run. No migrations, no new software your team has to learn.
Who actually does the work?
Me and my agents. Agents do the research, drafting, and monitoring. I do the judgment, the builds, and every client conversation. You will always know which is which.
What if we just want the plan, not the build?
The audit deliverable is written so any competent builder can execute it, me or anyone else. It is yours either way.
What makes this different from an automation agency?
Most automation agencies start with tools. I start with the operating problem: who owns it, where it stalls, which exceptions matter, and what a successful handoff looks like. Then I prove one system in your real stack before recommending a larger build.
Who is this not for?
Businesses whose model is paid media placement, PR distribution, or SEO services. If that is you, I will point you to someone good.