AI as infrastructure, not as gadget

Stop treating AI as a novelty. Start treating it as infrastructure for your practice. The five AI systems that senior consultants are deploying in 2026.

This is the edition I've wanted to write since day one. After 8 weeks covering pricing, networks, sales, and methodology, it's time to talk seriously about AI. No gadgets. No hype. The specific AI infrastructure that senior consultants can deploy this week to transform their productivity, delivery quality, and competitive advantage.

Francis Beaulieu

Why this matters right now

The tipping point is approaching. According to the 2025 Bain report on global consulting, 34% of consulting services buyers now expect AI-augmented delivery as standard. Not as a complement. As a baseline expectation. Within 12 months, that figure will exceed 50%. The consultant who hasn't integrated AI into their workflows will be perceived the way the consultant without email was perceived in 2005: functional but outdated.

The window to adopt is narrowing. The window to differentiate is still open.

Pricing: AI as a margin amplifier, not a price reducer

The action: When AI reduces your delivery time, do NOT reduce your fees. Instead, add a layer of deliverables that wasn't economically viable before: competitive analysis, scenario modelling, post-engagement tracking dashboard.

The logic: Tomasz Tunguz, investor and analyst at tomtunguz.com, documents that AI-native professional services firms charge premium rates because the quality of outputs exceeds what traditional delivery models produce, even though input costs are lower. Price for the output, not the input. The client is buying a result, not your hours.

The example: Your maturity diagnostic used to take 8 days and sold for $20,000. With AI, it takes 4 days but now includes a competitive benchmark, financial modelling, and a tracking dashboard. The deliverable is objectively better. The price goes to $25,000. Your margin doubles.

This week: Identify a recurring deliverable in your practice. Calculate the time AI saves you. Then identify what you could add with that recovered time. The result is a better deliverable at a better price.

Sales and business development: demonstrate AI in the sales conversation

The action: In your next discovery conversation, use AI live in front of the prospect to analyze an issue they just mentioned. Not as a show. As a demonstration of your actual workflow.

The script: "You just mentioned that your cost of non-quality increased by 15%. Let me show you how I'd start analyzing that." Open Claude. Feed in the context. Show the first set of hypotheses. Discuss the results together.

Why it closes engagements: The prospect doesn't see a promise. They see how you work. Process transparency builds trust. And live AI consistently impresses executives because most have never seen it used professionally.

This week: Prepare a 5-minute demonstration of your AI workflow for your next prospect. The first time you see a CEO lean toward the screen and say "wait, do that again," you'll understand why this is a powerful sales tool.

From the Field

Fatima, Victoria

I've been a digital strategy consultant for provincial and municipal governments for eight years. For the first two years of generative AI, I treated it as a glorified writing assistant. Document summaries, first drafts of reports, email rewording. Useful but not transformative.

Then I shifted my perspective. Instead of asking 'how can AI help me with what I already do?', I asked 'what services couldn't I offer before AI, but could I offer now?'

The answer was comparative policy analysis. Before AI, comparing 15 open data policies from 15 municipalities took 2-3 weeks. With AI, I do it in 2 days with a higher level of granularity. I launched a new service: 'Municipal Digital Policy Benchmarking.' Price: $18,000 per report. Production time: 3 days. No competitor offers it because nobody thinks to combine sector expertise with AI analysis at that scale.

My revenue increased by 40% in 8 months. Not by working more. By offering something new.

Collaboration networks: the AI learning group for consultants

The action: Assemble 4-6 senior consultants for a monthly AI learning group. Each participant shares one AI workflow they've deployed in their practice: what works, what doesn't, and the measurable results.

Why a group: Learning AI alone is slow. Individual experimentation is limited by your own imagination. A group of 5 consultants sharing their workflows gives you access to 5 times the experimentation. After 6 months, the group collectively owns a repertoire of 30+ tested workflows.

The format: 60 minutes monthly. Each participant gets 10 minutes: "Here's the workflow, here's the use case, here's the result, here's what I'd do differently." The last 10 minutes: open discussion on new capabilities or problems encountered.

This week: Send the invitation to 5 consultants you respect: "I'm starting an AI learning group for senior consultants. A 60-minute call per month where everyone shares a tested workflow. No selling, no competition. Collective learning. Interested?"

Value creation: the five AI systems of a modern practice

The action: Deploy these five AI systems in your practice. Not all this week. One per month.

System 1: Pre-engagement intelligence. Before every engagement, use AI to build an intelligence file on the client, their industry, and their likely issues. Result: you enter every engagement with hypotheses instead of blank questions.

System 2: Proprietary knowledge base. After every engagement, structure your findings, results, and lessons in a dedicated AI space (Claude Project, Notion AI). Result: after 20 engagements, you can query your entire body of experience.

System 3: Deliverable acceleration. Use AI for first drafts, data analysis, benchmarks, and financial models. Result: production time decreases; insight time increases.

System 4: Client intelligence monitoring. Set up a weekly AI-powered watch for your active clients. Result: you share proactive intelligence the client wasn't expecting.

System 5: Practice analytics. Quarterly, analyze your practice data (revenue, margins, satisfaction, pipeline) with AI. Result: strategic decisions informed by data instead of intuition.

This week: Choose the system that would have the most immediate impact on your practice. Deploy a minimal version. Iterate.

The AI Accelerator

Practice scenario: you're a cybersecurity consultant who performs risk assessments and compliance audits for SMBs.

1. Attack surface analysis. Feed the client's digital asset inventory (domains, subdomains, exposed applications, cloud services) into Claude. Ask: "Identify the most probable attack vectors by asset category, ranked by criticality and ease of exploitation." The AI produces a preliminary assessment that directs your technical analysis.

2. Multi-framework compliance assessment. Import applicable compliance requirements (Loi 25, SOC 2, ISO 27001, PCI DSS) and the client's existing controls. The AI creates a cross-compliance matrix that identifies controls satisfying multiple frameworks and common gaps. A single control can cover 3-4 requirements.

3. Security policy drafting. For each required policy (access management, data classification, incident response), feed in the organizational context and regulatory requirements. The AI generates a compliant first draft, tailored to the client's maturity level. You refine the specifics.

4. Incident scenario simulator. Ask the AI to simulate a security incident (ransomware, data exfiltration, credential compromise). "You are an attacker who has compromised a workstation. What lateral movements are possible based on this network architecture?" The exercise reveals weaknesses that static audits miss.

5. Executive report generator. Feed in your technical findings. The AI translates them into business risk language for the board: estimated financial impact per scenario, comparison with public incidents in the same sector, and recommendations prioritized by risk-to-investment ratio.

AI: the AI workflow that summarizes this entire edition

The action: Feed this edition into Claude. Ask: "Create a personalized AI implementation plan for a consultant in [your field]. Prioritize the 5 systems by impact and ease of deployment. For each system, provide the recommended tool, setup time, and first use case to test."

Why this meta-use matters: The best test of a tool is using it to improve its own use. If AI can take this edition's content and personalize it for your specific context, it just demonstrated exactly the type of value this edition is about.

Research by Ethan Mollick at Wharton, published on One Useful Thing, shows that professionals who personalize educational content with AI retain 40% more information and implement 2.3 times more actions, because personalization creates immediate relevance.

The process: (1) Copy the text of this edition. (2) Feed it into Claude with your profile: field, client type, practice size, current tools. (3) Ask for a personalized prioritization. (4) Start with System #1 from your personalized plan.

This week: Do it now. The plan AI creates is your AI roadmap for the next 3 months. If you only deploy one system, make it the proprietary knowledge base. That's the system that compounds most over time.

Like what you read? Get this in your inbox every Tuesday.

By subscribing, you agree to receive a weekly newsletter from Cogni6 inc. (Quebec, Canada). You can unsubscribe at any time.

Free. No spam. Unsubscribe in one click.