Yourdataiseverywhere.
Yourprocessesaremanual.
Youdon’tknowwheretostart.
No deck, no junior team, no six month roadmap. I walk into your operation, build what you actually need, and your team owns it from day one.
10 to 15%
Margin recovery
(construction)
8 hrs → 2 hrs
Grant writing time
(nonprofit consulting)
0% → 42%
Meeting booking rate
(legal services)
Work-Smart.ai is fractional AI leadership for mid-market companies. Ignacio Lopez works directly with CEOs and COOs to diagnose what is slowing the operation down and build the data layers, dashboards, and AI systems that fix it. Fixed-fee engagements, 4 to 16 weeks. Miami based. English and Spanish.
Your team runs on manual workarounds. Your employees are using ChatGPT without a policy. Your CEO can't get a straight answer without asking three people and waiting two days.
"Seven active projects and I find out we’re over budget a month after the fact. Cost tracking lives in Excel. Change orders get lost between the field and the office. The foreman reports by phone call."
Concreto. construction company, 650 employees.
"90% of our institutional knowledge is in Outlook. Proposals take 4 to 8 hours because we rebuild the same arguments from scratch every time. Nobody uses the tools we’re paying for."
Managing Partner. $14B wealth advisory firm, Miami.
"We know we need to do something with AI. We just don’t know what to do first. Everyone’s selling us a platform. Nobody’s telling us where to actually start."
CEO. professional services firm, 80 employees.
"Half my team is using ChatGPT on their personal accounts. I don’t know what they’re putting in there. If I need to worry about 10% inaccuracy, I don’t want to use it, but I can’t just ban it either."
Managing Partner. advisory firm, Miami.
Hire a consulting firm.
Big-firm pricing
They’ll send a team of six, run a 12-week discovery, and deliver a 200-page strategy deck. What you get: a document. What you don’t get: a working system. You need someone who builds.
Hire a full-time AI leader.
Full-time salary
Plus benefits, plus the 6-month search to find one. And in a market where the role is evolving every quarter, you’re betting a permanent headcount on a job description nobody can fully define yet.
Do it yourself.
Hidden costs
Buy some tools, assign it to your IT person or your most "tech-savvy" manager. This is how most mid-market companies end up with 15 disconnected tools, shadow AI on personal accounts, and an AI initiative that stalls in 90 days.
The fourth option:
Bring in someone who has already done this inside construction companies, family offices, law firms, manufacturers, and nonprofits. Fractional basis. Set the strategy. Build the system. Stay until it works. Leave when it no longer makes sense.
When I leave, the AI stays. Because you built it with me, not rented it from me.
Every company gets a different AI Operating System.
The logic is the same.
I walk into your operation, figure out where data is broken, where your team is losing hours, and where AI can make an actual difference. Then I build it. What that looks like depends on where you hurt most.
What I've built:
For a construction company losing 10 to 15% of margins on every project, and finding out 30 days too late:
The CEO would ask "how is project 3 doing?" and the answer took four days. Costs lived in a 10-tab Excel spreadsheet. Overruns happened in real time but nobody saw them until billing. I built Capataz, a system where the CEO types a question and gets an answer in seconds. Every project’s budget, schedule, and progress in one dashboard. Cost updates that used to take an hour of manual recalculation happen automatically. The company stopped discovering problems after the money was already gone.
650 employees. 9-month engagement.
For a $14B wealth advisory firm invisible to AI search, with six operational gaps nobody had mapped:
Started as a 6-week AI visibility project. The firm had more than a decade of published expertise but appeared in only 2 of 10 relevant AI queries. I extracted their Voice DNA, built 60 authority pages, and closed the visibility gap. Then the conversations with each partner revealed six operational problems: proposals written from scratch, knowledge trapped in email, tools deployed but not configured. What started as content became a 9-month Fractional Head of AI engagement. Proposal automation, knowledge systems, intelligence briefing, team training, and a full Operational Blueprint. Everything built belongs to the firm.
9-month Fractional Head of AI engagement.
For a law firm where the founding attorney was answering WhatsApp leads during billable hours:
Leads were coming in from ads. Nobody was converting them. The lawyers themselves were doing intake, asking the same qualifying questions over and over. What city are you in? What’s your budget? Is your company already formed? Meanwhile, there was no way to know how many leads came in last month or where they came from. I built Victoria, an AI assistant that answers every WhatsApp inquiry in seconds, asks only the questions that matter, and books qualified prospects directly onto the attorney’s calendar with a full briefing. I also rebuilt the website, set up lead tracking, and connected everything so the firm finally knows which ads are working and which leads turned into clients.
Live since August 2025.
For a nonprofit consultant spending 6 to 8 hours on every grant proposal:
She’d built a successful practice. $9M+ in grants secured across 20+ organizations. But her core deliverable, a grant application blueprint, took 6 to 8 hours every time. Her virtual assistant couldn’t replicate the work because the knowledge lived in the business owner’s head and scattered across six disconnected AI tools. I read everything she’d ever written, proposals, strategic plans, emails, workshop decks, and built a communication profile so the AI produces drafts that sound like her, not like a template. Then I structured her entire process into a shared workspace: grant blueprints, meeting debriefs, client emails, all running from the same system her VA accesses independently. Grant blueprints dropped from 6 to 8 hours to 2 to 3. Meeting debriefs from hours to minutes. Same team, same hours, 3 to 4x the output.
2 to 3 week setup. Joy of Impact.
You tell me what hurts.
30-minute conversation. You describe the operation. I listen for the data problems underneath. No pitch. No pressure. If there’s a fit, I’ll tell you. If there isn’t, I’ll tell you that too.
I diagnose.
Depending on the scope, this is a 2 to 4 week audit or a focused diagnostic. I map your data, your tools, your processes, your team’s AI usage. You get a clear picture, not a slide deck. A decision document.
I build.
Whatever the audit reveals, could be a data layer and dashboard, could be a WhatsApp AI agent, could be a full operational rebuild. Custom to your operation. Production-ready. First working system ships in 4 to 8 weeks. You own everything I build.
I stay (if it makes sense).
Monthly retainer for governance, optimization, new automations, team enablement. Not every company needs this. I’ll be honest about whether you do.
"Every buyer I talk to says some version of the same thing: we know AI matters, we just don't know where to start."
AI Ops Audit
Fixed-fee diagnostic · 2 to 4 weeks
Tax credits may offset 20-40% →You know something’s off but don’t know where to start. I map your data, your tools, your processes, and your team’s AI usage. You get a decision document, not a slide deck. The diagnostic that every engagement starts with.
For companies ready to stop guessing and get a clear picture.
Learn more →AI Foundation Build
Fixed-fee build · 4 to 16 weeks
Tax credits may offset 20-40% →You know what’s broken. I build the system that fixes it, dashboards, AI assistants, automations, data layer. Custom to your operation. First working system ships in 4 to 8 weeks. You own everything I build.
For companies ready to build, not just plan.
Learn more →Retained AI Operations
Monthly retainer
Tax credits may offset 20-40% →The system is running. I stay to govern, optimize, train your team, and build what’s next. Not every company needs this. I’ll be honest about whether you do.
For companies where AI is now part of how you operate.
Learn more →AI Training & Workshops
Per session · per session
Florida IWT grant reimburses 50-75% of training →Your team needs to use AI tools effectively. Hands-on, not lectures. Real workflows from their actual jobs. Every participant leaves with working prompts and systems they use the next day.
For companies that want their people productive with AI, fast.
Learn more →Over 60% of eligible companies never claim these credits. Don't be one of them.
The same R&D tax credit that pharmaceutical companies and tech giants have used for decades applies to custom AI consulting, and most mid-market companies don't know it. Every engagement I scope includes a funding qualification review, because across four programs, the typical offset is 20 to 30% of your total investment. For pass-through entities in Florida, it can reach 40%.
On a $25,000 AI investment
~35% offsetSection 174A
R&D Credit
IWT Grant
Net Cost
You pay: $16,270
Federal R&D Tax Credit (IRC §41)
When a company hires me to build a custom AI system, whether that is a dashboard, an AI agent, or an automation workflow, 65% of the consulting fee counts as a Qualified Research Expense. On a $25,000 AI Foundation Build, that generates approximately $975 to $2,275 in direct federal tax credits depending on your claiming history. The credit is dollar-for-dollar against your income tax liability. Construction companies, law firms, distributors, and financial services firms all qualify. You don't need to be in tech.
Section 174A Immediate Expensing (Restored 2025)
Congress permanently reversed the mandatory 5-year amortization of R&D expenses. Starting 2025, you expense 100% of domestic AI development costs in the year you pay them. For companies under $31M in revenue, this is retroactive to 2022, meaning if you paid for any custom software development or AI consulting in 2022, 2023, or 2024, you may be owed a refund.
The deadline: July 6, 2026. Amended returns for 2022 to 2024 must be filed by this date. If your company filed its 2022 return by April 15, 2023, the window may have already narrowed further. Talk to your CPA now.
Florida IWT Grant (Training)
Florida's Incumbent Worker Training program reimburses 50 to 75% of AI training costs, up to $100K per year. Companies with 50 or fewer employees get the 75% rate. A $5,000 AI workshop costs your company $1,250 out of pocket. I qualify as an approved training provider. The application is rolling, and the program is funded until exhausted.
SBA 7(a) Financing
For companies that want to spread the investment, SBA 7(a) loans explicitly list "AI-related expenses" as eligible. Up to $5M, with an 85% government guarantee on loans under $150K. This converts a $25K to $50K upfront investment into monthly installments, while you capture the tax credits and grants on the full amount.
| Your Investment | R&D Credit (Yr 3+) | Section 174A Value | IWT Grant | Total Offset |
|---|---|---|---|---|
| $25K build (construction, C-Corp) | $2,275 | $3,413 | • | $5,688 (23%) |
| $25K build + $5K training (≤50 employees) | $2,275 | $3,413 | $3,750 | $9,438 (31%) |
| $24K build + $5K training (pass-through, ≤50 employees) | $2,184 | $5,772 | $3,750 | $11,706 (40%) |
I walk every client through the qualifying criteria, connect them with R&D credit-experienced CPAs, and provide the documentation the IRS requires, project scope documents, iteration logs, technical narratives. This is part of every engagement, not an upsell.
See the full tax credits & grants guide with industry-specific qualification tables →Mid-Market Companies (Any Industry)
50 to 500 employees. $5M, $100M revenue. Too big for generic AI tools, too lean to hire a full-time AI team.
Your data lives in 5+ systems. Your team is already using ChatGPT without a policy. You know AI matters but you don’t have a roadmap. I come in as your fractional AI leader, diagnose, build, and stay until it works.
Take the free assessment →Financial Services & Advisory
Family offices, wealth advisors, asset managers.
Your data is in Outlook and your team’s institutional knowledge lives in people’s heads. Quarterly reports take weeks of manual reconciliation. Audit prep means reviewing hundreds of pages of regulatory documents by hand. I build the AI layer that automates reporting, makes 12 years of IP searchable in seconds, and makes your expertise visible to the AI systems your prospects use to find advisors.
See how AI Visibility works →Legal & Professional Services
Law firms, CPA firms, consulting firms. 20 to 500 employees.
Your attorneys bill 2.3 hours a day. The other 5.7 is admin, intake, and document work that AI handles better than another associate. I build the intake system, the document assistant, and the workflows that give you those hours back.
See what a fractional AI leader does →Construction
GCs, specialty contractors, developer-builders. 20 to 650 employees.
Your projects run on Excel, your capataz doesn’t report digitally, and you find out about cost overruns 30 days late. I build the system that gives you real-time visibility across every project.
See what I built for a construction company →Distribution, Wholesale & Manufacturing
Distributors, wholesalers, packaging, regional operators. 20 to 500 employees.
10 reports from 5 systems and nobody trusts the numbers. Your sales team carries physical catalogs. I build the data layer, the digital tools, and the dashboards that make your operation competitive.
Start with a diagnostic →You work with me directly.
That's the whole point.
I'm Ignacio Lopez. Computer Science degree, MBA, and years of building technology for operations teams. Eight client companies across construction, wealth advisory, legal, manufacturing, nonprofit, and distribution. From solo consultants cutting grant writing from 6-8 hours to 2-3, to construction companies tracking $50M in real time. First working system ships in 4 to 8 weeks.
I built Capataz for Concreto, a 650-person construction company that was losing 10 to 15% of margin because cost overruns went undetected for 30 days. Now the CEO asks a question about any project and gets an answer in seconds, not four days.
For a $14B wealth advisory firm, what started as a 6-week AI visibility project became a 9-month Fractional Head of AI engagement. I extracted their Voice DNA from more than a decade of published material, built 60 authority pages, then expanded into proposal automation, knowledge systems, intelligence briefing, and team training.
I rebuilt the website, lead tracking, and hiring portal for a packaging manufacturer operating across 4 countries. I built three websites, a client portal, and a project dashboard for Rudolph Architecture. I set up an AI workspace for Joy of Impact that cut grant proposals from 6 to 8 hours down to 1 to 2. I replaced 7 disconnected tools with one platform for Almaga in 3 months.
See the case studies →When you hire Work-Smart.ai you get me, plus every AI system I use to do the work of a team. Deep research agents that process hundreds of sources overnight. Custom AI assistants trained on your data. Voice DNA extraction that reads every document you have ever written. Automation that runs while nobody is watching. You are not hiring a solo consultant. You are hiring an operator who builds with the same tools he deploys for clients.
Based in Miami. Clients across the US and Latin America. Bilingual. English and Spanish.
Frequently Asked Questions
The AI Ops Audit is a 2 to 4 week diagnostic. The AI Foundation Build is a 4 to 16 week implementation. Retained AI Operations is an ongoing monthly Fractional Head of AI engagement. Every engagement is fixed-fee and scoped to outcomes, not billed by the hour. For exact pricing, see the AI consulting cost guide.
Any mid-market company with 50 to 500 employees. Financial services, legal, construction, distribution/wholesale, and professional services are where I have the deepest case studies. But the pattern, scattered data, manual processes, no AI roadmap, is the same regardless of vertical. If your data is a mess and your team doesn’t know where to start, we should talk.
First working system ships in 4 to 8 weeks. Not a slide deck. Not a prototype on a staging server. A production system your team uses on Monday morning. The AI Ops Audit (the diagnostic) delivers a decision document in 2 to 4 weeks.
Most of my clients have. The pattern is the same: someone sold them a tool without fixing the data underneath, or delivered a strategy without building anything. I start with the data layer. If your data isn’t ready, I fix that first. If the scope wasn’t grounded in real operational pain, that’s what the diagnostic is for. I don’t do pilots that sit on a shelf.
You work with me. There’s no team. No junior associate. No account manager in between. That’s the point of fractional, you get a senior operator at a fraction of the cost of a full-time hire, without the bait-and-switch of a consulting firm that sells the partner and delivers the analyst.
Where Should a Mid-Market Company Start with AI?
Most companies start with tools. That’s the wrong first step. Here’s what the ones that succeed do differently, and what an AI Ops Audit actually looks like from the inside.
Read more →How Much Does AI Consulting Cost for a Company with 20 to 500 Employees?
A short audit. A foundation build. An ongoing retainer. Here’s what drives the variation, what you should expect at each tier, and how to compare against a consulting firm or full-time hire.
Read more →How to Get Your Business Found by ChatGPT, Perplexity, and AI Search
Most citations come from URLs ranked outside Google’s top 10. The 7-step playbook to get cited when prospects ask an LLM for recommendations: answer capsules, schema, cross-platform presence.
Read more →Most of my clients started exactly where you are. They knew something needed to change. They didn't know what the first step should look like.
Start with a conversation. I'll be honest about whether I can help, and what makes sense for your situation.