Short answer
To choose an AI consulting firm, look for a partner that learns your business first, has technical people who write the code, can show deployed products, trains your team, and leaves you owning the code it builds. You can test all of that in your first conversations with the ten questions below. Lattice Partners is an AI implementation partner based in Los Angeles and New York that does AI strategy and team training and builds custom agents and software.
- Get a clear answer on who owns the code and what you'll pay after launch.
- Ask for a timeline to a first working version, in weeks.
What do AI consulting firms do?
AI consulting firms help companies decide where AI fits in their business and, depending on the firm, train employees and build the software. The work usually falls into three areas: AI strategy (finding and ranking use cases), training (teaching staff to use tools like ChatGPT and Claude), and implementation (building custom agents, integrations, and internal tools).
Firms differ in how far down that list they go. Some stop at a roadmap. A smaller group, often called AI builders or AI implementation partners, does all three with one team.
For a fuller picture, see what an AI consulting firm actually does.
What criteria should you use to choose an AI consulting partner?
When choosing an AI consulting partner, judge the firm on whether it will understand your business and whether your team can run what it ships once the engagement ends. These seven criteria cover most of that:
- Learns your business first. A good firm asks how work moves through your team and where it gets stuck before it proposes anything.
- Technical people who write code. At a smaller firm, look for technical founders. At a larger one, meet the engineering lead.
- Deployed products it can show. Live software with real users beats case-study PDFs.
- Training plus building. Adoption needs both. People ignore tools they weren't trained on, and training fades quickly when there's nothing new to use afterward.
- Clear code ownership. You should know whether you'll own the code or rent it.
- Honesty when AI isn't the answer. Some problems are better solved with a process change or an off-the-shelf tool.
- A short timeline to a first version. For a well-scoped problem, something should work in weeks.
What are the 10 questions to ask an AI consulting firm?
The ten questions to ask an AI consulting firm cover its process, its team, its track record, ownership, and timeline. Next to each one is what a good answer sounds like.
| # | Question | What a good answer sounds like |
|---|---|---|
| 1 | "How will you learn our business before recommending anything?" | Interviews with the people doing the work, a review of workflows and data, and a ranked list of problems. |
| 2 | "Who will write the code?" | Named people you can meet, ideally including founders or senior engineers who are on the sales call. |
| 3 | "Can you show us a product you've deployed?" | A live walkthrough of software in daily use, with client details hidden. |
| 4 | "Do you train our team, and on what?" | Hands-on training on the tools being built, plus general AI tools your team already has. |
| 5 | "Who owns the code when you're done?" | "You do." The software is a one-time build with no per-seat or per-module license from the firm. |
| 6 | "What would you tell us not to use AI for?" | Specific examples where a simpler fix beats AI, and a willingness to say so in your case. |
| 7 | "How long until we have a first working version?" | Weeks for a well-scoped first version, with a plan for what comes after. |
| 8 | "How will this connect to the systems we already use?" | Named integration approaches for your CRM, ERP (enterprise resource planning) system, or databases. |
| 9 | "What happens after launch?" | An iteration plan based on real usage. Most products need a few rounds. |
| 10 | "Have you worked with companies our size and in our industry?" | Relevant experience described generically, with workflows that sound like yours. |
Good answers name people and systems and give timelines in weeks. If a firm stays general across most of these, treat that as your answer.
Questions about deployed products and who writes the code are covered in more depth in how to tell if a firm ships software. What an AI builder is explains the difference between firms that advise and firms that build.
Why does code ownership matter when choosing an AI consulting firm?
Code ownership matters because it decides what you'll pay after launch and whether you can change vendors later. If a firm builds custom software and then licenses it back to you, you're paying rent on something built for your business.
ERP add-on pricing is a good example of the license model. NetSuite's own page says its CPQ (configure, price, quote) tool "is available as an add-on module," and that NetSuite licensing is an annual fee made up of the "core platform, optional modules and the number of users." That model is standard for platforms. For custom software, ask for ownership in writing and ask what ongoing costs, if any, remain after the build.
For timelines, see how long custom AI software takes to build.
What should mid-market companies look for in AI consulting?
Mid-market companies should look for AI consulting that fits a team without a large data science group and still delivers working software. These companies have real problems AI can solve but usually can't staff an internal AI team to carry a roadmap forward.
A roadmap is less useful when nobody in-house can build it. A mid-market company usually gets more from a business AI consulting partner that builds the first version and trains the people who'll use it, then stays involved while the product improves. For a metal fabrication shop, that first version might be a quoting assistant, while a PE (private equity) firm might start with a tool that screens companies against its mandate.
"Ask them to show you something they built that a client still uses every day. That one question tells you most of what you need to know."
Colin McDonnell, Co-Founder and Managing Partner, Lattice Partners
Once you've shortlisted firms, from AI strategy to production shows what a good handoff looks like. Why AI pilots stall shows what to guard against.
How Lattice Partners does this
Lattice Partners works with mid-market companies, often founder-led, in private equity and finance, manufacturing and metal fabrication, construction, architecture, media, and food service. Our founders, Colin McDonnell, Martin Anquetil, and Christian Yeghnazar, are technical and write the code themselves.
Engagements generally follow four steps:
- Learn. We study how your business runs and where time and money are wasted. If AI isn't the answer, we tell you.
- Tailor. We design around the data and systems you already have, at a level your team can work with.
- Build and ship. First versions of a well-scoped project usually ship within a few weeks, with hands-on training.
- Iterate. We improve from real usage. Most products go through two to three iterations.
Many clients start with training on tools like ChatGPT and Claude, then move into building custom agents and internal software. Some of that work is MCP (Model Context Protocol) integrations that connect AI assistants to CRMs, data sources, and internal databases. Cost depends on scope. Custom software is a one-time build that the client owns, with no recurring per-seat or per-module license from us. We can stay on after launch to support and improve what we built.
Frequently asked questions
How much does an AI consulting firm cost?
The cost of an AI consulting firm depends mostly on scope and on whether the engagement includes building software. A build that connects to several systems or a training program for a large team will cost more. Ask each firm to separate strategy, training, and build costs, and ask about ongoing fees after launch, since licensing models vary widely.
Should I hire a large consulting firm or a smaller AI firm?
A large consulting firm fits enterprise-wide programs with many business units and heavy governance needs. If you're a mid-market company that wants working software quickly, with direct access to the people building it, a smaller AI firm is often the better match. Ask either one who will be on your project day to day.
Do I need an AI strategy before hiring an AI consulting firm?
You don't need an AI strategy before hiring an AI consulting firm, because building one is part of what a good firm does. Bring a list of the workflows that frustrate your team most. A good firm will turn that into a ranked set of use cases.
What's a red flag when evaluating AI consulting firms?
A common red flag is a firm that recommends AI for every problem you describe. Other red flags are no deployed products to show and a recurring license on software built for you.
Related reading
About Lattice Partners
Lattice Partners is an AI implementation partner based in Los Angeles and New York. We help companies build an AI strategy, train their employees on AI tools like ChatGPT and Claude, and build custom AI agents and software. Clients own the code we build. Lattice Partners was founded by Colin McDonnell, Martin Anquetil, and Christian Yeghnazar. Contact us to talk about your use case.