AI consulting guide
The AI Consultant Roadmap: What to Actually Do in Your First 90 Days
By Bracken Fields
The AI consultant roadmap that works
If you're starting an AI consulting practice or thinking about it, you need a roadmap that helps clients see value fast. Your first 90 days should answer one question: can you help a client remove friction from one workflow and prove it was worth the cost?
This is the AI consultant roadmap I use. Fast iteration, measurable wins, no bullshit.
Days 1-30: Learn how to audit workflows before you pitch models
The first month is about developing a repeatable way to find high-value AI opportunities inside a business. You need a system for identifying where time and money leak out.
Week 1: Build your audit framework
Create a simple intake process to map recurring work. You're looking for workflows that repeat weekly, involve multiple people, touch clear inputs and outputs, and cost real money. Examples: sales intake, proposal generation, document review, customer onboarding, support replies, report generation, data entry.
For each workflow, capture: - Number of people involved - Weekly volume (hours, tickets, documents) - Systems used - Current pain points - Who owns the workflow
Week 2: Learn to estimate cost and value
Speak the language of business impact, not technical capability. Practice estimating the cost of workflows: hours per week × loaded labor rate + error cost + delay cost + opportunity cost.
A workflow costing $5,000/month in labor with measurable delays deserves more attention than a clever chatbot idea that saves five minutes.
Week 3: Identify your first real engagement
Find one business (could be your own, a friend's company, or a low-risk pilot client) and run a full workflow audit. Map 5-10 recurring workflows, estimate costs, and rank them by ROI potential.
The best first projects have: - High repetition and clear inputs - A named owner and low regulatory risk
Week 4: Build your service offering
Package what you learned into a clear service offering. Define your audit process, typical deliverables, timeline, and pricing. Write a one-page service description that focuses on business outcomes, not AI jargon.
By day 30, you should have a repeatable audit process and your first engagement lined up.
Days 31-60: Learn to build focused automations that people actually use
Month two is about execution. Pick one workflow from your audit and build something that works on Tuesday morning, not just in a demo.
Week 5: Master workflow design
The difference between AI that gets used and AI that gets ignored is workflow design. Learn to: - Map the current process exactly as it happens (not the idealized version) - Identify where AI can help (drafting, classifying, extracting, summarizing) - Define the human review step - Set clear acceptance criteria
Week 6: Build your first automation
Pick a specific tool or platform to master. Could be: - Make/Zapier for simple integrations - Python + OpenAI API for custom workflows - LangChain for more complex pipelines - Retool/Bubble for internal tools with UI
Build one focused automation: a Slack bot that answers team questions, an email pipeline that drafts responses, or a document processor that extracts structured data.
Week 7: Test with real users
Give your automation to the people who will actually use it. Watch where they get stuck, where they override the output, and where they avoid the tool entirely. Fix the workflow, not just the prompt.
Week 8: Measure and document
Compare the new workflow against your baseline estimate from month one. Measure: - Time saved per week and error reduction - Adoption rate and user satisfaction
Document what worked, what didn't, and what you'd do differently. This becomes your case study.
By day 60, you should have one working automation, real usage data, and a documented case study.
Days 61-90: Build your consulting practice infrastructure
The third month is about repeatability. Turn what you learned into systems you can reuse with every client.
Week 9: Create your service packages
Based on your first 60 days, define 3-4 core service offerings: - Workflow audit and AI roadmap ($2,500-$5,000, 1-2 weeks) - Focused automation build ($3,000-$8,000, 2-3 weeks) - Implementation and integration ($5,000-$15,000, 4-6 weeks) - Ongoing support retainer ($1,000-$3,000/month)
Week 10: Build your marketing assets
Create: - A clear website with specific services, pricing anchors, and case studies - 3-4 blog posts showing expertise (workflow audits, common AI mistakes, cost savings opportunities)
Week 11: Develop your sales process
Create a repeatable discovery and qualification process: - Initial call script (30 min): understand their business, identify pain points - Proposal template with specific deliverables, timeline, pricing
Week 12: Close your next two clients
Use your case study, service packages, and sales process to close two more engagements. One workflow audit (fast cash, builds relationship) and one automation build (demonstrates value).
By day 90, you should have three client engagements completed or in progress, documented service offerings, and a repeatable sales and delivery process.
Common mistakes on the AI consultant roadmap
Don't spend month one comparing model APIs. Spend it learning how to find high-value workflows in real businesses. The technical skill matters, but business judgment matters more.
Don't build a generic "AI assistant for everything." Build one focused automation that solves one specific problem completely. Depth beats breadth.
Don't wait until everything is perfect to charge real money. If you're delivering value, charge for it. Start with lower prices to build case studies, but don't work for free.
Don't ignore the human workflow. The best AI projects remove friction while keeping humans in control. If your automation bypasses necessary review steps, it won't get adopted.
What comes after the first 90 days
Once you've completed this roadmap, you have: - A repeatable audit and discovery process - Proven ability to build and ship focused automations - Real case studies with measured results - Clear service offerings and pricing - A pipeline of potential clients
From here, you can: - Scale by bringing on delivery partners or junior consultants - Specialize in a specific industry vertical (healthcare, legal, professional services)
The AI consultant roadmap is shipping value fast
Consulting means helping clients solve problems faster than they could solve them themselves.
Your first 90 days should prove you can: 1. Find high-value workflows 2. Build something that works 3. Measure the result 4. Charge money for it
Everything else is optimization.