AI Audit
Two to four weeks to find where AI will actually pay off before you spend on it: a ranked, costed plan, your current AI tools checked, and quick wins. If we can't find 300 hours a year of work to save, the audit is free.
See the AI Audit →Fractional Chief AI Officer
Off-the-shelf AI doesn't know your business. We build the one that does.
Your AI leader, without the full-time hire. We set the plan, build what's worth building and train your team, aiming for at least three times our fee back in measurable value. We review the numbers with you every quarter.
Operations overview · sample data
38New requests
27Quoted
11 minMedian reply
3Awaiting approval
RequestItemStatus
Northwind EquipmentExcavatorQuoted
Harbor Steel Co.Coil loadNeeds review
Summit BuildersCrane partsFollow-up drafted
Draft follow-up to Summit Builders is readyApprove
Most companies know AI matters. Few have anyone to own it.
The CEO is running the business. Operations is running the day. IT, if there is IT, is keeping the lights on. So AI ends up as a few chat subscriptions, a vendor pitch or two, and a sense that you're falling behind. Without someone accountable for it, the good ideas never get past the demo — and the risky ones slip in unnoticed.
Not sure where to start?
Two steps to find out where AI fits in your business.
- Step 1 · Free, about 5 minutes
Check your readiness
The AI Readiness Scorecard: 15 quick questions, then where you stand and the three things to do first.
Take the scorecardSee an example result - Step 2 · Your team, or the first part of the audit
Check the tools you already pay for
What can your current software's AI already do, and what's still missing for the work you want to change? Your team can work this out, or it's the first part of the AI Audit.
See the AI Audit
Use the results whether you work with us or not. Want a rough number first? Try the ROI calculator. Already have a proposal from an AI vendor? Check it with the free AI Vendor Check.
What we do
One accountable AI lead, starting with an audit.
Most companies are past “should we use AI?” Their people already use it, one chat window at a time. What's missing is the next step: putting AI into the real workflows of the business, with the access rules, approval steps, cost control and training that make it safe to rely on. That's the work I do, for owner-led companies that don't have an AI lead of their own, and without the enterprise price tag.
Fractional Chief AI Officer
A senior AI lead on a set monthly cadence. We build a plan you can defend to your board, pick and oversee vendors, and train your people. You get a short weekly check-in and a monthly results report.
How we work →How we recommend: first, switch on what you already pay for. Then off-the-shelf tools that do the job. We only build when nothing else fits, and when we do, we tell you in writing, score it the same way, and show you the off-the-shelf prices next to ours.
Offers
Three ways to start.
Pricing is set after the discovery call, based on your size and scope.
AI Audit
Find out where AI will actually pay off before you spend on it: a ranked, costed plan, your AI tools checked, and quick wins. If we can't find 300 hours a year of work to save, the audit is free.
See the AI AuditFractional Chief AI Officer
Your part-time AI lead. Strategy, vendor choices, training, a weekly check-in and a monthly results report. Small improvements are included; larger builds are quoted separately so you always know what you're paying for.
Book an AI discovery callAI Foundation 90
A defined first project: discovery, Hollis setup, and your first two or three workflows live, with a results review at day 90. A good fit if you want to see results before committing to a monthly role.
Book an AI discovery callA typical first year
What the first year usually looks like.
- Months 1–2
Discovery
We learn how the work really gets done and agree the first priorities.
- Months 2–4
First automations
One or two workflows go live, each with a person approving the output.
- Months 3–5
Platform setup
Hollis is connected to your tools, starting read-only.
- Months 5–9
Expand
We add the next workflows by return, one at a time.
- Every quarter
Results review
Months 3, 6, 9 and 12: what paid off, what didn't, and what's next.
Every company is different; this is a typical path, not a fixed schedule.
First projects
Where it usually starts.
First projects are small, measurable and close to revenue. In logistics, that often means:
- SalesDrafting quotes from emailed requests, and following up on leads nobody called back.
- OperationsTurning order emails into entries in your system, and flagging missing details.
- Carrier and contract checksComparing rate confirmations and contracts against what was agreed.
- BillingCollections reminders and invoice follow-ups, drafted for a person to send.
- Quality controlAn AI check on anything before it reaches a customer.
Why a fractional lead
A one-off AI project vs. a fractional AI lead.
| Typical AI project | Fractional AI lead | |
|---|---|---|
| Focus | One tool or use case | Your whole operation, by priority |
| After launch | Handed over, often fades | Measured, tuned and expanded |
| Who's accountable | The vendor, until the invoice | One named person, month to month |
| Vendors | Often recommends its own product | Neutral; picks what fits |
| Your team | A training session at the end | Ongoing coaching and simple guides |
| Results | Reported once | Monthly report + quarterly review |
Good fit
We're usually a good fit if…
- You don't have an in-house AI or IT lead, and no one owns “our AI plan”.
- Your leadership or board has started asking what you're doing about AI.
- You've seen AI demos that never made it into daily work.
- Your team re-types the same information between email, spreadsheets and your main system.
- You can name one person inside who will champion the work with us.
How we work
How we work with you.
Discovery
We spend two to four weeks learning how your work actually gets done: interviews by role, walking through each workflow, and looking at where time goes. You get a ranked list of opportunities, each with the cost, the risk and how we'd measure it.
Monthly results report
Every month you see what changed: hours saved, cycle times, what shipped, what's next. If something isn't paying off, the report says so.
Weekly check-in
A short standing call to keep priorities straight and decisions moving.
Human approval, always
Anything sent to a customer, anything that spends money, anything that deletes data: a named person on your team approves it first.
AS
Who you'll work with
Alexis Sockwell, founder.
Before AI, I spent more than 23 years as an orthopedic Physician Assistant in high-volume surgical practice. That work runs on discipline: checklists, careful documentation, patient privacy under HIPAA, and a second check before anything that can't be undone. I bring the same habits to AI.
Today I build and run AI systems myself, including a freight quoting system in production and Indelio, a quality management system for FDA-regulated companies. So when I recommend a tool, or recommend against building one, it comes from having done the work.
Wake Forest University: B.S. in Health and Exercise Science, and the Physician Assistant Program at Wake Forest School of Medicine (2001). Four-year letterman on the football team.
Examples
Things we've built and run ourselves.
We build our own tools so we know what it takes. We only recommend them when they score best for you.
VoxCoreHaul
- The problem
- Oversize freight quotes depend on dimensions, weight, permits and escorts: rules that usually live in a few experienced people's heads.
- What we built
- A quoting system that works out the equipment, permits and escorts for a request and prices it, with a person approving anything unusual.
- What it shows
- Built and proven in production with a heavy-haul brokerage: real requests, real loads, real pricing rules. It's where our habits come from: measure the outcome and keep a person in the loop.
Hollis by Voxelio
- The problem
- Managers and reps spend their day hunting for answers across email, spreadsheets and their main system.
- What we built
- An assistant connected to a company's own tools. It answers questions and drafts follow-ups and summaries, and a person approves anything before it's sent.
- What it shows
- How AI can sit on top of the tools you already have, with your data in your own accounts. Each part is marked with where it stands today.
Indelio
- The problem
- Quality teams at small medical-device and biotech companies run the same controls as large ones, often with one or two people and a stack of spreadsheets.
- What we built
- An AI-assisted quality management system: controlled documents, CAPA, training and more in one place, with electronic signatures and a tamper-evident audit trail.
- What it shows
- AI in a regulated setting: the AI drafts and suggests, and a person always reviews and signs.
Who it's for
Who it's for.
Growing companies — typically $5–100M in revenue — that don't have an in-house AI lead, especially in:
- Logistics and transportation
- Construction and the trades
- Manufacturing and equipment
If your operation runs on rules, judgment and steps only experienced people know, that's where we're most useful. If AI isn't the right answer yet, we'll tell you.
Voxelio AI was founded by Alexis Sockwell and is operated by Reshawn Holdings LLC in Sarasota, Florida. We work with a small number of companies at a time so each one gets real attention.
The discovery call
What happens on the call.
30 minutes on Zoom. No pitch deck.
- First 10 minHow your operation runs today and where time goes.
- Next 10 minTwo or three places AI could realistically help, and what it would take.
- Last 10 minWhether we're a fit, and the next step if we are. If we're not, we'll say so.
Resources
Resources.
The AI Discovery Playbook
The field guide we use in the first weeks with a new company: the questions we ask by role, the workflows we watch, how we rank what we find, and the one-page recommendation template.
Read the playbookAI for logistics
Where AI pays off for brokers, carriers and heavy-haul operators, with examples and the guardrails we use.
Read the guideAI Vendor Check
Score an AI vendor or agency proposal before you sign it: an 11-part scorecard, red flags, and the questions to send the vendor. Runs in your own Claude account.
Get the free toolDecision Board
Run a big decision past six reviewers built for your industry (freight, construction & trades, or any business) and get a clear verdict before you commit.
Get the free toolAI Readiness Scorecard
Fifteen short questions, a score for five areas, and the three things to do first before you buy or build anything. Your score on screen right away, no sign-up.
Take the scorecardMore articles coming.
FAQ
Questions we hear.
What does a Fractional Chief AI Officer actually do?
Owns AI for your company part-time: strategy, priorities, vendor choices, build oversight, training, and a monthly report on results. Like a fractional CFO, but for AI.
What does it cost?
It depends on size and scope. We set the fee so the expected value is at least three times higher, and we show you the numbers every month. You'll get a straight figure after the discovery call.
Is our data safe?
Your data stays in your own accounts wherever possible, access is limited to what each task needs, and every action is logged. Our own systems are reviewed against external security scans. Nothing goes to a customer, spends money or deletes data without a person approving it.
Our IT team will say no.
We'd rather work with them from day one. We document what connects to what, use your existing accounts and permissions, and IT can switch any piece off.
Our data is a mess.
Most companies' data is. We start where the data is good enough and clean up as we go, not the other way round.
We tried AI and it fizzled.
Usually because no one owned it after the demo. Ownership is the job here: a named lead, a monthly report and a quarterly review.
Is this about cutting jobs?
Our focus is taking repetitive work off your people so they can spend time on customers and judgment calls. How you use the time saved is your decision.
How do you measure whether AI is working?
In dollars and hours, each tied to the line of your P&L it's meant to move: revenue, cost of delivery or operating expenses. We take a baseline before anything changes and measure again after, using your own numbers. You see the results in the monthly report.
What if it doesn't pay off?
You'll see it in the monthly report before anyone else does. We stop or change what isn't working, and the quarterly review is where we decide together whether to continue.
Will you try to sell us your own software?
No. We recommend what you already pay for first, then off-the-shelf tools. We only suggest something we've built when it scores best for you, and we show you the off-the-shelf prices next to ours.
Do we need technical staff?
No. We handle the technical side and explain it in plain language.
Will customers know they're dealing with AI?
Yes. We disclose AI use where it touches your customers.
Let's see if it's a fit.
Tell us how your operation runs. We'll tell you honestly whether AI is worth it for you right now, and where we'd start.
Book an AI discovery call