Your support
team doesn't have an engineer inside it yet.
I'm a founder who built and sold an ecommerce business. Now I embed with ops teams, find what's worth automating, and ship production systems your team owns.
Cancel rate
▼ 1.2pp4.1%
On-time
▲ 0.4pp96.3%
Vendor avg
▲ 0.3B+
Vendor performance
7d
Live triage
Pinnacle Dist. cancel rate 14.1%
SP-1042: 6 stockouts past 4h
WH-2 East cycle trending +2.1d
01 · Go deep
Sit inside the data, the ops, the team
Tickets, vendor feeds, P&L, the people doing the work. No assumptions until I’ve seen it from the inside.
02 · Find the signal in the noise
Separate what actually moves the number
Most ops chaos is a signal-to-noise problem. I figure out which thread to pull before pulling anything.
03 · Action the next best thing
Ship the highest-leverage fix, not the biggest one
Built into your stack, automated where it can be, handed back clean. No retainer required, but you’ll never want me to go.
AI Automation & Agents
Production-grade AI automation, built by an operator who understands the business it serves.
AI agents. Agentic workflows. Retool apps. API integrations. Most ops consultants can tell you what to automate. I actually build it, deploy it, and hand it off. That combination of business context and build ability is genuinely rare in ecommerce.
AI-Powered Ticket Triage
Automatically classifies, routes, and prioritizes support tickets using AI, so nothing falls through and agents work the right things first.
- Handles roughly one in four monthly tickets end-to-end with no human touch
- AI classifies ticket type, urgency, and owner on intake
- Auto-routes to correct queue based on vendor or issue pattern
- SLA countdown visible per ticket, and escalation fires automatically
Agent Management & SLA Monitoring
A real-time ops dashboard that tracks agent workload, ticket SLAs, and escalation status. Built in Retool, owned by the team.
- Live view of every open ticket by agent, status, and SLA health
- Automated escalation when tickets approach SLA breach
- Workload balancing signals prevent queue pile-ups
“Fast shipping, filter fit perfectly. Only wish the packaging was...”
“Thank you! We're looking into packaging improvements — feedback passed to our ops team.”
Automated Review Processing
AI reads incoming Google reviews, generates on-brand responses, and routes edge cases for human review, without a person touching every one.
- AI categorizes each review by sentiment and topic
- Generates a tailored, on-brand response draft
- Negative or complex reviews flagged for human before posting
Proof
Real outcomes. No hedging.
These aren't case studies dressed up in consulting language. They're things that happened, stated plainly.
AI support automation
~25%
of tickets fully automated
Built AI-powered triage that classifies, routes, and resolves a significant share of monthly support contacts end-to-end, with no human touch required.
Ops automation build
A full role's weekly work → one click
A recurring weekly reporting process that consumed an entire role was reduced to a single button press. Built into the existing stack, documented, and handed off.
Catalog Automation
200,000+
SKUs added through automation
Built the automated catalog pipeline for an e-commerce store I owned and operated, scaling the product catalog past 200K SKUs without manual listing. The store was later sold.
Inventory & catalog diagnostics
Thousands of hidden SKUs unlocked
Found a single inventory feed bug silently suppressing a large portion of sellable catalog. Fixed the root cause, restored visibility, and dollarized the impact for leadership.
Cross-silo diagnostics
Cross-silo root cause found
A returns spike that looked like a product quality issue turned out to be a catalog data problem feeding bad specs to customers. Traced it across three teams, fixed the source, and the symptom disappeared.
Why a person, not a platform
Building automation is easy now. Knowing what to automate isn't.
Getting it adopted in a company with real people, politics, and unwritten rules is harder still.
That takes 15+ years in ecommerce, including running my own business, plus hands-on building. It's the part the AI tools you're being sold don't come with.
So I sit inside your team, learn how the work really gets done, and build automation your people actually use.
The problem nobody's naming
Most organizations operate with internal bleeding that goes undetected.
The business looks like it's running fine. Orders ship, dashboards refresh, the team stays busy. Underneath, margin is leaking through workflows the org stopped questioning years ago: too normalized to count as a problem, too distributed to land on anyone's roadmap.
Finding the bleed is how I decide what's worth automating first.
BLEED · 01
Vendor cancel rates everyone learned to live with
Cancellations get re-categorized as “out of stock” and stop being a metric. Margin walks out the door every day and nobody flags it because the dashboard still looks green.
BLEED · 02
A spreadsheet someone runs every morning
It started as a temporary fix two years ago. Now an entire function depends on it, and nobody knows what happens when that person takes a vacation.
BLEED · 03
Catalog data the team has stopped trusting
Specs are wrong, taxonomy is duplicated, technical product details don't line up across the catalog. Support and returns quietly absorb the cost, ticket by ticket, and it never traces back to the source.
BLEED · 04
Reports that take three people and a Tuesday
The number is right by Wednesday afternoon. By then the decision has already been made on instinct. The org runs on lagging signal and nobody calls it a problem.
Who I'm built for
Mid-market ecommerce ops teams. More manual work than hands.
You run vendor ops, fulfillment, catalog, or support for a growing ecommerce company. The team is stretched, the manual work keeps piling up, and there's no engineer in the room. If that sounds like you, keep reading.
Vendor relationships are unmanaged and it's costing you
Fill rates are bad, cancel rates are climbing, and nobody knows which vendors to hold accountable. I instrument vendor performance and build the systems that fix it.
Drop ship fulfillment is manual and breaking SLAs
Someone is hand-picking vendors, chasing statuses, updating spreadsheets. I build automated routing logic that picks the right vendor on every order, with no human in the loop.
Inventory and catalog data is a mess
Feeds are inconsistent, SKUs are duplicated, taxonomy is broken. I normalize the data layer so catalog scales to hundreds of thousands of products without adding headcount.
You know you need automation but don't know where to start
The team runs the same manual workflows every day. I identify the highest-leverage processes and build production-grade automations your team actually uses.
Services
Start with what to automate. Then build it.
Most teams start with the Roadmap, move into a build sprint, then keep me on as their fractional automation lead. Engagements are fixed-scope with clear deliverables, priced against the outcome, and you own everything at the end.
2–3 weeks
Automation Roadmap
What to automate, ranked by dollar impact. A deep dive across vendor, fulfillment, catalog, and support operations to find the work that is worth building first.
- Full ops diagnostic across all layers
- Data analysis: fill rates, cancel rates, SLA compliance
- Automation candidates ranked by dollar impact, with ROI estimates
- Written roadmap you can act on, with or without me
You get a ranked list of what to automate and a plan to build it. Not a generic framework, a map of your actual system.
4–8 weeks
Embedded Build Sprint
I sit with your team and ship. Ticket triage, SLA monitoring, agentic reporting, review automation, built in your stack and deployed alongside the people who will run it.
- Workflow design with the people doing the work
- Build in Retool, n8n, or your existing stack + AI APIs
- Full deployment, testing, and iteration
- Team training and complete documentation
A production-grade automation your team owns and uses every day, with me on hand as it grows. No vendor lock-in. Most teams keep me embedded to build the next one.
Ongoing · Monthly
Fractional Automation Lead
I build your first automation systems, then stay embedded as your fractional automation lead, setting the roadmap and building alongside your team.
- First automation systems built and deployed in your stack
- Train your team to run what's built, and help you hire when you're ready
- Documented systems your team owns, with no black boxes
- A monthly automation roadmap and build capacity as the function grows
You get working automation, a team that owns it, and an automation lead in the room who keeps it shipping.
3–6 weeks
Operational Triage & Stabilization
For when things are breaking. I come in, map what's actually happening, instrument the data, remove blockers, and hand you a working system.
- Rapid system mapping: what's breaking and why
- Data instrumentation so you can see what's actually happening
- Blocker removal: vendor issues, routing failures, catalog gaps
- Documented processes your team owns
You get a stabilized operation and a team that owns it. Not a slide deck, a working system.
Need a different scope or ongoing engagement?
Retainer, advisory, white-label partnerships, and one-off operator calls available. Let's talk.
Sean Dawes
Founder, Operator, Builder
About
Founder. Operator. Builder.
When I walk into an ops team, I'm not pattern matching from a client engagement or a case study. I'm recognizing something I've lived through. That's the thing that makes trust happen fast and keeps clients coming back.
I've actually done what
you're trying to do.
Built it. Scaled it. Sold it. If your ecommerce operations need someone who's been in the seat... not just advising from the side... let's talk. I'll tell you what I see and whether I can help.