Across the Table Build. Grow. Thrive.

AI & Technology Industry Stack 20 · Professional service firm Free Business Insights

The Professional Services Stack

Nine systems for engineering, architecture, advisory and IT firms — where utilization is the business and AI changes it.

About 7 min read

Download Knowledge Brief (PDF) .pdf · 151 KB

Who this is for

You run a firm that sells expert work by the project or the engagement: engineering, architecture, business advisory, IT consulting, technical services. You have several people, a pipeline, and a delivery process that has to be profitable per engagement.

This guide is for the firm leader who has outgrown the solo consultant’s toolkit and needs the pipeline, the delivery and the numbers to connect.

If you are still a practice of one, start with the consultant stack and come back when you hire.

Start with the work, not with the apps

These are not twenty lists of AI apps. Each stack starts with the jobs a particular business performs, then matches AI and conventional software to those jobs. Treat it as a menu to evaluate, not a shopping list.

For this kind of business the work runs in one line, start to finish: Lead → Engagement → Delivery → Billing → Relationship

Broken into the steps a customer actually moves through:

Lead → CRM → Discovery → AI meeting intelligence → Proposal → Project management → AI-assisted delivery → Invoice → Analytics → Renew

Read that line left to right. A lead becomes an engagement, the engagement is delivered, it is billed, and a good firm turns it into the next one rather than starting over.

A professional services firm runs on one number: utilization — the share of your people’s time that is billable. Every other metric is downstream of it. What makes this moment interesting is that AI changes both sides of the equation: it reduces the hours a deliverable takes, and it makes the non-billable work cheaper too.

The stack, piece by piece

Every tool below is matched to one of those steps. You do not need all of them, and you should not buy all of them at once. Read the “What it is” column first and find the step where your business is losing the most time or money.

ToolThe step it coversWhat it isPrice
HubSpotCRMCustomer records, marketing, sales and service in one platform.Free CRM; paid hubs per month.
CalendlyMeetingsShared booking links tied to your calendar.Free tier; paid per user/month.
Fireflies / FathomMeeting intelligenceRecords, transcribes and summarizes meetings, then pushes notes into your other tools.Free tier; paid per user/month. / Generous free tier; paid per user/month.
PandaDocProposalsProposals, quotes and contracts with e-signature.Per user/month.
Asana / ClickUpDeliveryProjects, tasks and workloads with AI summaries.Free tier; paid per user/month.
NotionKnowledgeDocs, wikis and databases with AI search across them.Free tier; paid per user/month.
ChatGPT / ClaudeAnalysis and draftingOpenAI’s general assistant — writing, analysis, research and, with Agent mode, multi-step work.Free tier; paid plans per user/month.
QuickBooksFinanceThe small-business accounting standard — books, invoicing, reporting.Per month by plan.
Power BIPerformance analyticsMicrosoft’s business intelligence and dashboards.Per user/month.

Where two names share a row with a slash between them, they do the same job and you pick one, not both. The description shows the first of the two.

This is a full stack and no firm should implement it in one quarter. CRM, meeting intelligence and project delivery first; analytics once there is enough clean data to analyze.

What to watch out for in each one

None of these are reasons not to buy. They are the things owners tell us they wish they had known first.

HubSpot — Costs step up sharply by tier and contact count.
Calendly — Set buffers or your day disappears.
Fireflies — Tell people they’re being recorded. Set a retention rule.
Fathom — Fewer integrations than the bigger platforms.
PandaDoc — Template setup takes an afternoon.
Asana — Structure only works if everyone updates it.
ClickUp — Configuration can become the project.
Notion — Becomes a junk drawer without structure.
ChatGPT — It sounds confident when it’s wrong. Check anything that touches money, law or a customer.
Claude — Same rule as any assistant: verify facts and never paste data you wouldn’t email.
QuickBooks — Only as clean as your categorization habits.
Power BI — Data modeling is the real work.

Where AI actually fits in this business

Research, drafting, meeting intelligence, deliverable assembly, utilization and performance analysis.

In plain terms, that means:

Capturing every client meeting as a transcript and a summary, so decisions do not live in one person’s notes.
Assembling the first version of a deliverable from your firm’s own past work and methodology.
Doing the background research on a client’s industry fast enough to be useful before the first meeting.
Drafting proposals from a structure instead of from scratch.
Reading utilization and project margin to show which engagements actually made money.

The firm-level version of the consultant’s problem: if you bill hours and your hours drop, your revenue drops with them. Firms that handle this well move toward fixed-fee or value-based engagements, where the efficiency lands in margin. That is a leadership decision and it should be made deliberately rather than discovered at year end.

Who this covers

Engineering consultants, architects, business advisory firms, IT consultants and other knowledge-based businesses.

The second-order effect is about junior staff. A great deal of what juniors used to do — first drafts, research, document assembly — is exactly what AI does well. Firms still need to develop people, so the development has to be designed in deliberately rather than happening as a by-product of grunt work.

What to set up first

Do these in order. Finish one before you start the next. Most owners who try to do all of it in one month end up using none of it.

  1. Get the pipeline into one CRM. Not a spreadsheet and not several people’s inboxes. You cannot forecast capacity against a pipeline you cannot see.
  2. Record and summarize client meetings. With consent. This is the cheapest institutional memory a firm can buy, and it stops decisions from living with whoever attended.
  3. Run delivery in one project system. Every engagement, with owners and dates. Utilization reporting is impossible without it.
  4. Build the knowledge base. Your methodology, your templates, your past work, searchable. This is what makes AI assistance actually useful, because it has your material to work from instead of generic material.
  5. Then add analytics. Utilization, realization and project margin. Once the first four are clean, these numbers become trustworthy, and they will change how you price.

What it costs to start

Nearly everything here is per user per month, so your stack cost tracks headcount. That is predictable and it also means reviewing seats when roles change.

The analytics layer can be the most expensive piece and should be the last one you buy. It is worthless on messy data and valuable on clean data, so sequence accordingly.

The decision that dwarfs all of this is pricing model. A firm that improves efficiency by a third and keeps billing by the hour has given its clients a discount it never agreed to.

The price column in the table above tells you how each one charges — per user, per location, per contact, or a percentage of every sale — not what you will pay. That is deliberate. The charging model is what decides whether a tool stays affordable as you grow, and it is the part that does not change every quarter. Get the current dollar figure from the vendor’s own pricing page before you budget.

The rules underneath all of it

Legal · Privacy · Data protection · Security · Compliance · Records · Audit · Risk · Human oversight — underneath every one of the twenty stacks, not an afterthought.

Client confidentiality governs tool selection. Know, for each vendor, where data is stored, who can see it and whether it trains models.

If your work carries professional licensure — engineering, architecture — the stamp and the professional judgment behind it cannot be delegated to software, and that should be stated in your AI policy.

Recording meetings requires consent, and in some states consent from everyone present.

Client contracts increasingly address AI use. Read what you are agreeing to, and have a firm position ready for clients who ask.

Written AI policy, communicated to staff, before tools are adopted — not after someone has already pasted a client document somewhere.

Mistakes that cost the most

Improving delivery efficiency while billing hourly, and shrinking your own revenue.
Keeping the methodology in senior people’s heads, which caps growth and makes every departure expensive.
Buying analytics before the underlying project data is clean.
Adopting AI tools with no written policy and finding out afterwards how they were used.
Removing the junior work without replacing the junior development that came with it.

The CEO decision

What is your actual utilization rate, and do you trust the number?
Which engagements last year made money, and which ones only looked like they did?
If delivery hours dropped by a third, would your revenue rise or fall?
Is your methodology written down, or carried by individuals?
How will your juniors learn the judgment that AI is now doing the first draft of?

Your quick review

☐ Every opportunity lives in one CRM.
☐ Client meetings are recorded and summarized, with consent.
☐ All engagements run in one project system.
☐ The firm’s methodology is written down and searchable.
☐ There is a written AI policy staff have read.

Additional resources

HubSpot — www.hubspot.com — Small businesses that want one system of record.
Calendly — calendly.com — Ending the back-and-forth on meeting times.
Fireflies / Fathom — fireflies.ai / fathom.video — Teams who keep losing decisions made on calls.
PandaDoc — www.pandadoc.com — Sales documents that need to look good and get signed.
Asana / ClickUp — asana.com / clickup.com — Teams who need visible ownership and dates.
Notion — www.notion.com — Your company knowledge in one searchable place.
ChatGPT / Claude — chatgpt.com / claude.ai — Your default thinking partner when you want one tool that does most jobs well.
QuickBooks — quickbooks.intuit.com — Clean books your accountant already knows how to read.
Power BI — www.microsoft.com/en-us/power-platform/products/power-bi — Microsoft-centered reporting.
Across the Table AI Assessment — acrossthetable.biz/#assessment

Go deeper on the pieces

Each of these is a full guide comparing the three tools that do that job:

Run the work — covers Notion, ClickUp, Asana
Ask and reason — covers ChatGPT, Claude
Capture meetings — covers Fireflies, Fathom
Fill the pipeline — covers HubSpot
Manage the money — covers QuickBooks

Continue learning

The Across the Table AI Assessment takes fifteen minutes and starts from the work the firm performs.

The guides above compare the leading tools for each job side by side.

Read and connect with Across the Table

Across the Table website: https://acrossthetable.biz/

The Scaling Edge newsletter: https://acrossthetablebiz.substack.com/

All Knowledge Articles