ThriveAI vs the Big Four and large consultancies
You can hire a Big Four firm such as Deloitte, EY or PwC, or a consultancy such as Accenture, to build AI into your business. You can also hire ThriveAI. This comparison is for owners, presidents and operations leads at smaller manufacturers and distributors in Ontario and Quebec. It sets out what each firm says publicly about who it serves, how its teams work, what it publishes about price and where your data sits. It also says when a large firm is the better choice.
The short answer
- Pick ThriveAI if you run a smaller manufacturer or distributor and want a working tool on your own data for the one job that costs you the most. Stage one is a working prototype for that job, at a fixed price, in weeks, and you keep it if you stop there.
- Pick a Big Four firm or Accenture if you need many engineers at once, if the AI work is part of a company-wide SAP, ServiceNow or Microsoft rollout, or if you need a supplier with an audited security certification. Accenture built its mid-market business, Accenture Edge, for companies with US$300 million to US$3 billion in annual revenue.
- Use both if a large firm already advises you and you want one job fixed on your own data while that work continues. If that firm is also your auditor, read the section on audit independence first.
Who you are comparing
The Big Four
The Big Four are the four largest audit firms. Canada’s audit regulator, the Canadian Public Accountability Board, reports on them as a group in its 2025 annual report, and together they audit 92.5% of Canadian reporting issuers by market capitalization. This page compares ThriveAI with three of them, Deloitte, EY (Ernst & Young) and PwC (PricewaterhouseCoopers), using what each firm publishes on its own pages.
Accenture
Accenture is a consultancy and does not audit financial statements, so it sits outside the Big Four. It appears on this page because it sells the same kind of AI build work. Its third-quarter fiscal 2026 results report about 799,000 people, about 9,000 clients and 104 client bookings of US$100 million or more in the fiscal year to date.
ThriveAI
ThriveAI is an AI engineering company in Ottawa that builds private AI systems for manufacturers and distributors in Ontario and Quebec, on their own data. It works hands on with your team. Derik Lawlis founded ThriveAI in December 2025, leads every project and stays close to the build. Before ThriveAI he worked in sales at American Iron & Metal, co-founded Mindbliss, which was later acquired, worked as a product manager at Invert, and built and sold Paddle.ca in May 2025. For more information, see About ThriveAI.
Every firm on this page sells build work
On September 23, 2026, every large firm on this page offered AI build work on its own pages: Deloitte and Accenture through forward deployed engineering, PwC through its agent OS platform and EY through agent build as a service. Forward deployed engineering means engineers who work inside the client’s business and build on its own systems.
Working with Derik to build these AI agents has been fun, very worthwhile, with a huge ROI for my team.
Fahd Alhattab, Founder, Unicorn Labs.
Side-by-side comparison
Each row states what the firms publish on their own pages, checked on September 23, 2026. Where the pages say nothing, the table says so.
| Question | ThriveAI | Big Four and Accenture |
|---|---|---|
| Who it serves | Smaller manufacturers and distributors in Ontario and Quebec. | Deloitte’s and Accenture’s alliances with Anthropic focus on regulated industries. Accenture Edge serves companies with US$300 million to US$3 billion in revenue. |
| Size | A small company in Ottawa, founded December 2025. | Accenture has about 799,000 people. Deloitte gives more than 470,000 people access to Claude. |
| Who does the build | Works on site with your team. Derik Lawlis, the founder, leads every project. | Deloitte describes “pods of specialist professionals” embedded with your team. Pod size is not published. |
| First thing you get | A working prototype for one job, at a fixed price, in weeks. You keep it if you stop. | Accenture Edge offers pre-built industry agents with Google Cloud and says clients can deploy solutions in weeks. |
| Published rates or minimums | None. Every project is quoted on its own. | None on any page checked, including Deloitte’s. |
| Platforms | Built on your own data, with the model you choose. | Built through alliances, for example Accenture with Microsoft, ServiceNow and SAP, and Deloitte with Anthropic. |
| Where your data sits | The platform is designed to keep your data on its own server in Canada. A hosted model may process requests outside Canada. | No default location for AI work stated on any page checked. |
| Your ERP | Read as it is, including older versions installed on your own server. | PwC says you can run pilots while you upgrade ERP or cloud systems. |
| Language | Working sessions in French or English. | Not stated on the pages checked. |
| If the firm audits you | ThriveAI does not audit financial statements. | If one of the Big Four audits you, independence rules limit what it can build for you. |
| Audited security certification | None. ThriveAI holds no SOC 2 or ISO 27001 certification. | Ask for the certificate that covers the team doing your work. |
What each firm sells for AI work
The table sets out what each firm publishes about its AI build work, checked on September 23, 2026. None of these pages states a price, a minimum or a default data location for AI work.
| Firm | What it sells | Alliances | Who it targets |
|---|---|---|---|
| Deloitte | Forward deployed engineering, with small pods embedded with your team. The service page gives no pod size. | Anthropic, with Claude available to more than 470,000 Deloitte people. In Canada, Zora AI, built on NVIDIA AI, and alliances including AWS, Google, SAP and ServiceNow. | Regulated industries, under the Anthropic alliance. |
| Accenture | Forward deployed engineering with Microsoft, ServiceNow and SAP, each on that vendor’s platform. Accenture Edge gives mid-market clients one team that scopes and builds the work, with one engagement lead. | Anthropic, through a business group that will train about 30,000 Accenture professionals on Claude. Edge works with Google Cloud and AWS. | Edge serves companies with US$300 million to US$3 billion in annual revenue. |
| PwC | Agent OS, a platform that coordinates AI agents from different vendors. PwC says you can run pilots while you upgrade ERP or cloud systems. | Anthropic, with Claude Code and Cowork rolling out, starting with its US teams. PwC Canada lists alliances from Microsoft to AWS. | PwC Canada’s AI page describes no offer for mid-sized companies. Its new finance group built on Claude starts with banking, insurance and healthcare. |
| EY | The ey.ai Agentic Platform and four service models, including agent build as a service. | None stated on EY Canada’s agentic AI page. | Not stated. |
A Deloitte US job posting for a lead engineer describes pods of two to five engineers, anchored onshore and supported offshore, with a posted salary of US$189,200 to US$372,900. The posting lists US cities only, so it does not describe Deloitte Canada.
On its third-quarter fiscal 2026 call, Accenture said mid-market clients do not need the coverage model it uses for larger enterprises, and that at least one in every two advanced AI projects leads to a data project. If your records are spread across an ERP, email and spreadsheets, plan for that data work whichever firm you hire. Accenture’s Canadian site has no Edge page of its own, and Edge’s AWS release names 407 ETR in the Greater Toronto Area as a case example.
A new entrant for mid-sized companies
On May 4, 2026, Anthropic, Blackstone, Hellman & Friedman and Goldman Sachs announced a new AI services company for mid-sized companies. It names mid-sized manufacturers among its targets, and Anthropic Applied AI engineers work alongside its teams. It is not one of the Big Four, and it belongs on your list if you are comparing larger providers.
Where your data sits
The platform ThriveAI builds on is designed to keep each client’s data on its own server in Canada. You choose the model that reads it: one that runs on that server, or a hosted model under a written agreement that the provider keeps nothing after answering. A hosted model may process requests outside Canada, so the contract names the model and its service tier.
None of the large-firm pages checked for this comparison states where your data sits by default for AI work. Deloitte Canada and SaskTel announced an alliance on September 23, 2025 for a planned AI factory and Applied AI lab in SaskTel’s Saskatchewan data centres. The release describes planned capabilities and does not describe a service in operation.
Senior Canadian leaders say this matters. In a Deloitte Canada survey of 300 senior leaders, most from companies with more than $100 million in revenue, 96% said it is important that AI is developed, hosted and governed under Canadian laws. Ask every firm, ThriveAI included, to name in the contract where your data is stored and which model processes it.
What it costs
None of the Big Four or Accenture publishes hourly rates, rate cards, engagement minimums, seat minimums or standard contract terms for AI work on the pages checked on September 23, 2026. The Deloitte forward deployed engineering page is typical: it describes the teams and publishes no price. Any figure you read elsewhere for a large-firm AI engagement is an estimate, so ask each firm for a written quote.
ThriveAI publishes no rates either. Every project is planned and quoted on its own, and stage one is a working prototype for one job at a fixed price.
Government of Canada Job Bank gives a reference point for what the engineers themselves earn. It lists a median wage for software engineers of $56.73 an hour in Ontario and $53.60 an hour in Quebec. For data scientists, the medians are $47.69 an hour in Ontario and $46.00 an hour in Quebec. All four figures come from Statistics Canada’s Labour Force Survey for 2023 to 2024. They are wages, and a firm’s billed rate also covers its overhead and margin.
Questions to ask any firm before you sign
- Who works on your project, and how many hours does each person spend on it?
- Where do the engineers work? Deloitte’s US posting describes pods anchored onshore and supported offshore.
- Is there a minimum engagement size or a minimum number of seats?
- What do you keep if you stop after the first stage: the code, the data, the prototype?
- Where is your data stored, which model processes it, and does the model provider keep anything after answering?
- If the firm audits your financial statements, which parts of the work do independence rules forbid?
When your auditor offers to build it
If one of the Big Four audits your financial statements, check the independence rules before it builds anything that touches your books. Under CPA Ontario Code Rule 204.4(28), a firm that audits your statements cannot design or implement a financial IT system that produces information significant to those statements, or that forms a significant part of your internal controls over them. The exception applies only if your company meets all five conditions in the rule, including that it makes all management decisions, evaluates the system and operates it.
The rule is stricter for reporting issuers and listed entities. It applies to audit firms, so it does not limit Accenture or ThriveAI. If your auditor is in Quebec, ask it which CPA Québec rule applies to your case.
When to pick each
Pick ThriveAI when
- You run a smaller manufacturer or distributor in Ontario or Quebec, and a few experienced people hold up order entry, purchasing, scheduling or quoting.
- You want to see a working tool on your own data before you commit to more. Stage one is a working prototype for one job, at a fixed price, in weeks, and you keep it if you stop there.
- You want to keep the ERP you run, including an older version installed on your own server.
- You want your data kept on its own server in Canada, with the model that reads it named in the contract.
- You want working sessions on site with your team, in French or English, with the founder leading the project.
- You want someone on your team to approve every draft before anything is sent or saved in your systems.
For examples of what these tools do at each desk in a plant, see AI for manufacturing, desk by desk.
Pick a Big Four firm or Accenture when
- Your company is at the scale these firms build for. Accenture Edge starts at US$300 million in annual revenue, and the Deloitte and Accenture alliances with Anthropic focus on regulated industries.
- You are rolling out SAP, ServiceNow or Microsoft across the company and want the platform’s partner to lead the AI work. Accenture runs forward deployed engineering programs with all three.
- You need many engineers at once, across several sites or countries.
- You want AI work tied to finance advice from the same firm, and that firm does not audit you. PwC’s new Office of the CFO business group focuses on client finance organizations and is built on Claude.
- Your customers or lenders require a supplier with an audited security certification such as SOC 2 or ISO 27001. ThriveAI holds neither.
- Your board wants a large firm’s name on the work.
Use both
A large firm can run strategy, risk work or an ERP upgrade while ThriveAI builds one tool on your own data. The systems ThriveAI builds read your ERP and other software and change nothing in them, so they can run next to other projects. Each firm signs its own agreement with you.