I've spent months figuring out how to teach this: a scalable framework to learn how to multiply YOUR expertise with AI. This is the same mental model I use to 20x my productivity with AI!
Nate
I keep getting asked to share how I get so much done.
I keep getting asked to share how I’ve 20x’d (or more) my ability to generate quality work in 20 years.
But I thought I’d do one better than that: I want to explicitly lay out an easy-to-use framework for scaling expertise using AI across more than 20 different job roles.
I’m laying out for you my exact mental framework for leveraging AI to scale my own expertise, so you understand how Nate gets so much done.
But beyond that, I want you to be able to grab these prompts and use them to start scaling your own expertise. Because that’s what this really is: a universal framework for scaling expertise.
AI gives us tools to scale expertise that just weren’t possible even two years ago.
And the thing is, learning this framework now sets you up for the 50x+ scaling multiple I expect in the next year. AI will keep getting better, and that means free additional multiples on expertise next year for people who figure this out now.
We all have different kinds of expertise, so I’ve included a REALLY wide range here.
Here are 27 prompts to get you going on scaling expertise in a way that works for you.
Tech and product Prompts
- Senior Software Engineer — RFC Document
- Vibe Coder — Technical Spec
- Product Manager — PRD
- Engineer — Technical Design Doc
- CTO — Engineering Strategy Memo
Go-to-market and customer-facing prompts
- Marketing Consultant — Campaign Strategy
- Head of Sales — Sales Strategy Document
- Entrepreneur — Investor Update
- Customer Success Manager — Account Plan
- Insurance Agent — Policy Comparison
- Amazon FBA Seller — Product Description
Finance and legal prompts
- CPA — Tax Strategy Letter
- Financial Advisor — Quarterly Portfolio Report
- Attorney — Discovery Responses
Healthcare prompts
- Family Physician — SOAP Notes
- Dentist — Treatment Plan Letter
- Veterinarian — Surgery Recommendation
Built environment and home services prompts
- Architect — Project Proposal
- General Contractor — Remodel Estimate
- HVAC Contractor — Estimate
- Plumber — Repiping Estimate
- Electrician — Panel Upgrade Proposal
- Landscaper — Backyard Renovation Proposal
- Real Estate Agent — Property Listing
Personal services and events prompts
- Wedding Planner — Client Proposal
- College Counselor — Student Recommendation Letter
YES, you might notice the expertise doesn’t stop at tech roles! Expertise is a universal human skill, and AI can help you scale even if you don’t work in tech. That’s part of why I spent time crafting so many prompts.
All this and we’re not done yet! On top of all 27 prompts, I’ve put together:
- A 5 minute TLDR to explain how it all works
- A complete how-to guide
- Common pitfalls and what to watch for
- A clear framework you can scale to other areas of expertise
Basically, I wrote the guide I would have wanted when I started to figure out how to scale with AI by trial-and-error a couple of years ago. I wrote the guide to help you understand the keys to how I think, and how I construct prompts to scale my own expertise.
By the end of this guide, you won’t just have 27 prompts—you’ll be able to use this framework to write any prompt you need to scale your own expertise. You’ll be ready to scale your expertise past 50x as AI gets better and better over the next year.
Subscribers get all these newsletters!
Subscribed
Grab those 27 prompts now!
Here’s what you’re getting: 27 prompts across every domain where expertise hits a documentation bottleneck. These matter because seeing the pattern once isn’t enough—you need to see it applied across different fields to understand what stays the same and what changes.
The consistent structure across all 27 prompts is the teaching method: every single one follows the same four-part formula (your role, your audience, your goal, your constraints), then you add your raw expertise.
The structure never changes, only the domain-specific application changes. The goal here is to get you practicing in a very consistent way so you can see the specific spots you need to edit very quickly once the prompt results come back.
Once you see this pattern repeat across legal work, medical documentation, trade estimates, and engineering specs, you understand how to build your own prompts for whatever you do. Don’t read all 27—search the document for your field and start there, then look at a few others to see what transfers.
The TLDR
The constraint on expertise has never been your knowledge. It’s the translation layer. You can diagnose a problem in minutes, but documenting that diagnosis takes hours. You know the strategy immediately, but writing the brief takes all day. Your brain works fast. Documentation works slow. That ratio is the bottleneck.
AI removes it—but not in the way most people think. Large language models are pattern-matching engines trained on billions of examples of professional documentation. They’ve seen thousands of legal briefs, medical charts, project proposals, and estimates. They know the structure, the format, the professional conventions. What they don’t have is your specific expertise, your client context, your professional judgment. That’s the gap.
And that’s why bad prompting makes for bad expertise, and why people give up.
The breakthrough is this: when you provide the expertise within a useful, predictable prompt to the LLM, LLMs can leverage that expertise to build a professional output in a fraction of the time.
You give it four things—your role, your audience, your goal, and your constraints—plus your raw knowledge in whatever form is fastest for you. Voice notes, bullet points, technical observations.
The LLM maps your expertise onto the professional formats it knows, handling the translation from “I know this” to “here’s a polished deliverable.” You review for accuracy, adjust for nuance, ensure quality. The model gets you 80% there. You handle the 20% that requires judgment.
This gets better as models improve, and this gets better as you practice more.
Today’s LLMs are fully capable of getting to 80%. Tomorrow’s will understand domain-specific nuance, maintain consistency across documents, and require less verification—they’ll get to 85% maybe. But the pattern stays the same: you own the expertise, the model owns the documentation bottleneck.
And I want to emphasize one thing here—there’s a scaling limit on LLM ability to replace expertise. They just do not have your years of experience or problem specific knowledge. So you need to work with them in such a way that you can predictably touch the work and bring your expertise to bear—the thing that makes it great.
This article shows you exactly how it works with 20 domain-specific examples, walks through the four principles that make it work, and gives you the five-step process to implement it in your field. Whether you’re in law, medicine, trades, consulting, or engineering—if your expertise is the product, this is how you scale it.
How to Scale Your Expertise with AI
=====================================
For thousands of years, there have been only three ways to scale your expertise. All of them suck. AI just invented a fourth way, and most professionals still don’t know about it.
Here’s the problem that nobody talks about: expertise is the one thing in business that doesn’t scale naturally. You can scale products by manufacturing more. You can scale content by publishing more. You can scale distribution to reach more people. But expertise only lives in your brain, and there have been exactly three ways to work around this constraint.
The Three Traditional Ways to Scale (And Why They Fail)
Option one: work more hours. You’re a lawyer, client demand goes up, so you work nights and weekends until you burn out. This doesn’t scale because there’s not infinite time in the day.
Option two: hire people. You bring on associates, junior lawyers, apprentices—whoever. But they don’t scale expertise, they dilute it. The junior person isn’t you. They don’t have the years of pattern recognition that come with real expertise. Every piece of work they touch needs review by someone who actually knows what they’re doing. You end up trading your actual work for management work, and you’re still the bottleneck.
Option three: raise your prices. You charge $600 an hour, then $1,000 an hour. But there’s a ceiling. Eventually you’re too expensive for most clients, and you’ve traded volume for rate while still being limited by time.
These are your options. This is why every expertise-based business hits a wall. Whether you’re in legal, medicine, senior engineering, HVAC installation, plumbing, or any other field where knowledge is the product, your expertise hits a scaling wall. Your knowledge is the asset, and there’s only one of you.
Enter AI and the fourth way. But the breakthrough isn’t what most people think it is.
The Real Constraint Wasn’t Your Expertise
The thing nobody says out loud: the constraint has never really been your expertise. It’s been the translation layer.
Look at an HVAC contractor diagnosing a failing system. That takes her 20 minutes. She knows what’s wrong—undersized unit, leaking ductwork, whatever it is. She has 15 years of experience. The diagnosis is fast.
Writing the estimate that explains this to a homeowner takes 2.5 hours. It has to be professionally formatted, translated into accessible language, explain why this solution and not the cheaper one, include photos with annotations, be persuasive enough to win the job. Her expertise took 20 minutes. Documenting her expertise took 2.5 hours.
That ratio has been the problem. That’s been the bottleneck. That’s what doesn’t scale.
And this is true everywhere. A senior attorney knows the legal strategy in minutes but takes hours to write the brief. Even with a paralegal, getting the intent down and getting it polished takes forever. The doctor knows the diagnosis but completing the chart note takes time. The architect knows the design solution but creating the presentation takes much, much longer.
Your brain works fast. Documentation works slow.
The fourth way—the AI way—attacks that bottleneck directly. It separates your expertise from the documentation of that expertise. And once you see the pattern, you recognize it everywhere.
The Universal Pattern: Sarah’s Story
Let me walk you through how this actually works with one complete example. Once you see the pattern here, you’ll recognize how it applies to whatever domain you work in.
Sarah runs a residential HVAC business in the suburbs. She’s excellent at her work—15 years of experience, knows the equipment cold, can diagnose a struggling system in 20 minutes flat. Her constraint was never expertise. It was estimates.
Here’s what her old workflow looked like. She’d do a site visit and diagnosis in 45 minutes, then spend 2.5 hours writing the estimate. That meant an opening letter explaining the problem, a line-item breakdown with justifications for each cost, an explanation of why this solution instead of cheaper alternatives, photos with annotations showing what she found, and an explanation of financing options.
Total time per estimate: over 3 hours. She could write maybe 5 estimates per week, which meant 15 hours of writing time. With a 35% close rate, she was winning about 1.75 jobs per week.
Sarah is an expert, not a writer. But homeowners need professional, detailed estimates that build trust and explain technical work in accessible language. The translation from “I know exactly what this house needs” to “here’s a document that wins the job” took longer than the diagnosis itself.
Here’s what her new workflow looks like. The site visit and diagnosis still take 45 minutes—that hasn’t changed and shouldn’t change, because that’s where her expertise matters.
But now, after the inspection, she records a 5-minute voice memo. “Three-bedroom ranch, current unit is 2-ton from 2008, struggling to cool second floor in summer. System is undersized for square footage, ductwork looks original and has leaks in the attic runs. Recommend 3-ton high-efficiency unit, ductwork sealing and two additional returns for better air circulation. Will improve comfort and cut their energy bills by about 30%. House has good insulation so this should solve it.”
Then she spends 2 minutes giving AI a prompt. “I’m an HVAC contractor. Turn these notes into a professional estimate for a homeowner. Explain why the upgrade is necessary without using technical jargon. Emphasize comfort improvements and energy savings. Include a section on why proper sizing matters. Tone should be reassuring and educational, not salesy.”
Then she drives to her next site visit while AI works. She reviews the output in about 8 minutes, adding her pricing for labor and materials and timeline, inserting her photos, adjusting any technical details AI got slightly wrong, and tweaking the tone if needed. Final review and send takes another 5 minutes.
Total time: 20 minutes of writing work.
Now she writes 25 estimates per week instead of 5. That’s 8 hours of writing instead of 15, which frees up 7 hours. Her close rate stayed at 35%—same quality, same trust-building, just faster. But 25 estimates at 35% means 8.75 won jobs per week instead of 1.75. She 5x’d her opportunity pipeline by removing the documentation bottleneck.
Here’s what matters in this example: AI didn’t replace her expertise.
The diagnosis is hers—that system is undersized, the ductwork has problems, these are professional judgments earned over 15 years. The solution is hers—3-ton unit, ductwork sealing, additional returns, this is technical expertise. The pricing is hers—she knows her costs, her market, her margins. The relationship is hers—this is still Sarah talking to her customer, building trust through transparency. And quality control is hers—she reviews every estimate, catches technical errors, ensures accuracy.
What AI handles is the paralegal work, the apprentice work, the junior associate work. The structure: professional format, logical flow, proper sections. The translation: technical knowledge into homeowner language. The persuasion: why this matters, why not the cheaper option, what the benefits are. The consistency: every estimate is polished, thorough, professional.
Sarah’s expertise stopped being trapped by the documentation bottleneck. That’s the breakthrough, and that’s the pattern that works everywhere.
The Same Pattern Across Every Domain
Once you see the pattern, you recognize it in every field where expertise is the product. The bottleneck is always the same: translating expertise into deliverables. The workflow is always the same: you provide judgment and facts, AI provides structure and polish. What changes is the domain-specific application.
Legal Work
A commercial litigation attorney spends 6 hours drafting responses to interrogatories. She knows the facts, knows the strategy, knows what to say and what to protect. But translating “we deny this, we admit that, here’s why” into properly formatted, citable, formal legal documents takes forever.
Now she spends 30 minutes outlining her responses—what to admit, what to deny, what objections to raise, what facts to cite—gives AI the structure and tone requirements, reviews the output for accuracy and privilege issues, and sends it in 90 minutes total.
What she still owns: legal strategy, privilege decisions, factual accuracy, risk assessment, client interests. This is the practice of law. What AI handles: citation format, formal legal language, document structure, consistent tone across 50 interrogatory responses.
Discovery responses that used to take all day Friday now take Friday morning. She handles 4x more matters without compromising quality, bills for expertise instead of document formatting.
Medical Work
A family physician sees patients all morning, then spends two hours after clinic completing charts. She knows the diagnosis, knows the treatment plan, knows what happened in the appointment. But translating “patient presents with X, diagnosed Y, prescribed Z” into proper SOAP notes, insurance-friendly language, and complete documentation steals hours from patient care.
During or immediately after each appointment, she records brief bullet points. AI structures these into complete documentation following her template and clinic standards. She reviews for clinical accuracy and signs off. Post-clinic documentation drops from 2 hours to 20 minutes.
What she still owns: diagnosis, treatment decisions, clinical judgment, patient safety, prescription decisions. This is the practice of medicine. What AI handles: formatting notes into proper structure, translating shorthand into complete sentences, ensuring documentation meets insurance requirements, maintaining consistent template structure.
A physician who saw 20 patients per day now sees 24—not because appointments are shorter, but because documentation no longer creates a ceiling on capacity. Or she keeps the same patient load and leaves on time instead of staying late to finish charts.
Ecommerce
An Amazon FBA seller spots a trending product category but can’t move fast because learning the category, researching competitors, writing 20 optimized product descriptions, and understanding customer language takes weeks. By the time she’s ready to list, the opportunity window might be closing.
Now she researches the top products in a category in 2 hours, identifies gaps and opportunities, sources products from her existing suppliers, then gives AI her product specs plus competitive research. She has 20 listings written and optimized in a day and tests the category while it’s still hot.
What she still owns: product selection, pricing strategy, supplier relationships, inventory management, competitive positioning. What AI handles: product descriptions, keyword optimization, A+ content, comparison tables, benefit-focused copy, multiple variations.
She used to operate in 2-3 categories she knew deeply. Now she tests 10+ categories per quarter. Most don’t work, but the winners more than cover the failures. The game changed from “expertise in one niche” to “rapid testing across niches.” Her advantage is velocity.
Real Estate
An agent can only serve 8 clients simultaneously because every listing requires multiple deliverables—property description, social media posts, email campaigns, CMA narratives, open house materials, buyer follow-ups. She has the market knowledge and relationship skills to serve 20 clients, but content creation is the ceiling.
When she gets a new listing, she takes photos and notes key details in 15 minutes. AI generates the listing description, social posts, email announcement, and open house materials simultaneously. What used to take 3 hours per listing takes 30 minutes.
What she still owns: property valuation, negotiation, client relationships, market knowledge. What AI handles: listing copy, social media content, email campaigns, market update newsletters, CMA narratives, follow-up sequences.
She serves 20+ active clients simultaneously. Her close rate stays high because the quality of materials stayed high—she just removed the time constraint. More clients, same expertise, same service level. Revenue tripled without hiring an assistant.
The pattern holds everywhere. Your expertise is the valuable part. The documentation layer was the constraint. AI removed it.
Four Principles That Make This Work
These principles work whether you’re in tech or not in tech, whether you’re practicing law or installing systems or writing code. Once you understand them, you can apply this pattern to your own domain.
First principle: expertise compounds, documentation doesn’t. You get better at your craft every year. You see patterns faster. You make decisions with more confidence. This is why business studies show expertise peaking much later in career arcs than people realize—into your 50s and beyond. But writing still takes the same amount of time.
AI makes documentation compound with your expertise. That’s the breakthrough. Your knowledge keeps growing, and now your output can keep pace with it.
Second principle: quality control lives with you. The lawyer still reviews for legal accuracy and privilege issues. The doctor checks the diagnosis and treatment appropriateness. You don’t outsource your judgment. You outsource the translation. AI can draft, but you’re still the professional responsible for what gets delivered.
Third principle: the 80-20 threshold. AI will get you 80% of the way much, much faster than any other method. Faster than a paralegal, faster than a nurse taking notes, faster than any apprentice. But that last 20% requires your expertise, and that’s exactly where you want to be spending your time.
You want to get hands-on with your business in the right way. You need to set up the prompt and the context to make sure AI delivers the right 80%, then you apply your expertise to the remaining 20% that actually requires your judgment.
Fourth principle: context is your multiplier. This is the secret to making everything else work. The better you can articulate what you need, the better the output. If you just say “write an estimate,” it’s not going to be great. If you just say “write a generic NDA,” it’s not going to be great.
Your ability to articulate what you need in structured, templatized context forms is your superpower. You want to be able to say: this is what I need, this is the context you need for this task, this is the context you need about me, this is the context you need about the client or the patient or whoever the reader is, and this is my expectation for this draft specifically. The more clear you can be, the more likely you are to get the right 80% and have only the important 20% left to touch.
That volume you get to creates more optionality for you.
And that’s the real payoff. When documentation was the bottleneck, you turned down work. You could only scale your time so far. By attacking this bottleneck, you unlock optionality because your expertise is no longer the constraint. That’s the 10x return that AI delivers that none of the older scaling levers could touch.
How to Actually Do This
The framework is universal. The implementation is domain-specific. Here’s how to start.
Step One: Pick Your Bottleneck
Identify the one repetitive task where you spend hours translating expertise into deliverables. Common candidates:
- Discovery responses and client letters in legal work
- Patient documentation and insurance letters in medical work
- Estimates and change orders in trades
- Product descriptions and customer service responses in ecommerce
- Listings and CMAs in real estate
- Design docs and RFCs in engineering
- Client reports and proposals in consulting
Choose the task you do weekly that takes 2+ hours but doesn’t require your highest-level expertise to document.
Step Two: Learn the Context Formula
AI needs four things to produce professional output: your role, your audience, your goal, and your constraints.
For a family law attorney, that might look like: “I’m a family law attorney drafting a response to discovery interrogatories. The audience is opposing counsel and potentially a judge. I need to answer these questions while protecting privilege and avoiding admissions. Tone should be formal and cautious. Here are my notes on each interrogatory.”
For a family physician: “I’m a family physician documenting a patient visit. This needs to follow SOAP note format for insurance purposes. The patient is a 45-year-old with Type 2 diabetes presenting for follow-up. Here are my visit notes. Make this complete and insurance-friendly while staying clinically accurate.”
For a general contractor: “I’m a general contractor writing an estimate for a kitchen remodel. The homeowner is price-sensitive but values quality. I need to explain why certain choices—structural work, permits, quality materials—are necessary and not just upsells. Tone should be educational and reassuring. Here’s what the job requires.”
The pattern is always the same: establish who you are, who you’re writing for, what you’re trying to accomplish, and what constraints matter. The better you get at articulating this context, the better AI performs.
Step Three: Use the Iteration Pattern
Professional output rarely comes from one prompt. Think of this as a conversation, not a command. You get the first draft from your initial prompt. Then you refine:
- Make the second paragraph more specific about timeline
- This sounds too salesy, make it more educational
- Include a line about code compliance requirements
- Shorten this by 20% without losing key points
You’re collaborating with a very fast assistant who needs direction. Give it.
Step Four: Know What to Verify
Every domain has critical elements that you must check yourself. AI can draft, but you’re still the professional.
In legal work, check:
- Factual accuracy—AI cannot know your case specifics
- Privilege issues—nothing protected should be revealed
- Jurisdiction-specific rules—statutes, case law, procedures
- Strategy decisions—what to admit, what to contest
In medical work, check:
- Diagnosis accuracy
- Treatment appropriateness
- Drug interactions and dosages
- Patient-specific history—AI doesn’t have full context
In trades, check:
- Code compliance
- Safety requirements
- Pricing accuracy
- Site-specific conditions—you saw it, AI didn’t
In ecommerce, check:
- Product specifications
- Verify claims
- Competitive accuracy
- Brand voice
In real estate, check:
- Square footage and property details
- Verify comparable sales data
- Confirm disclosure requirements
- Avoid fair housing violations
In engineering, check:
- Technical accuracy
- Security implications
- Performance claims
- Operational concerns
The principle is simple: AI handles the translation layer, but the expertise layer is yours. Never outsource what requires professional judgment.
Step Five: Start Small, Then Scale
Weeks 1-2: Use AI for one task only. Get comfortable with the workflow. Learn how to prompt for your domain.
Weeks 3-4: Add a second task type. Start seeing the time savings compound.
Month 2: Expand to all appropriate documentation tasks and calculate your actual time savings.
Month 3: Reinvest the freed time. Take on more clients, improve service quality, or just leave on time.
The goal isn’t to use AI for everything. It’s to remove the bottleneck between your expertise and your deliverables.
Your Domain Here
Don’t see your field in the examples above? The pattern works everywhere. Here’s how to map it to what you do.
Accounting or tax preparation. The bottleneck is client reports and tax strategy explanations. The unlock is turning numbers into narrative in minutes. You still own tax strategy, compliance, and calculation accuracy. AI handles translating P&L into client insights, explanation letters, and year-end summaries.
Architecture or engineering. The bottleneck is project documentation and client presentations. The unlock is turning technical specifications into persuasive proposals. You still own design decisions, code compliance, and structural calculations. AI handles RFP responses, client-facing narratives, and translating technical details into accessible language.
Education or training. The bottleneck is lesson plans, assessments, and differentiated materials. The unlock is taking one concept and creating multiple formats and levels instantly. You still own pedagogical expertise, student assessment, and curriculum goals. AI handles creating variations, formatting materials, and parent communications.
Non-profit or fundraising. The bottleneck is grant applications and donor communications. The unlock is turning mission into compelling narratives at scale. You still own program expertise, impact measurement, and donor relationships. AI handles tailoring stories to different funders, application formatting, and appeal letters.
Consulting in any domain. The bottleneck is client deliverables and proposals. The unlock is turning strategic insights into polished documents. You still own client diagnosis, recommendations, and expertise. AI handles structuring slide decks, writing reports, and formatting frameworks.
Hospitality or events. The bottleneck is event documentation and vendor coordination. The unlock is turning logistics into professional communications. You still own creative vision, vendor relationships, and day-of execution. AI handles proposals, timelines, guest communications, vendor briefs, and post-event summaries.
Financial advisory. The bottleneck is client investment reports and explanations. The unlock is turning portfolio data into personalized insights. You still own investment strategy, risk assessment, and fiduciary responsibility. AI handles market commentary, portfolio narratives, and client education materials.
Insurance in any type. The bottleneck is policy explanations and claims documentation. The unlock is turning coverage into client-understandable explanations. You still own risk assessment, policy recommendations, and claims decisions. AI handles coverage summaries, claims letters, and policy comparison documents.
See the pattern? First, identify your bottleneck—what takes time but doesn’t require your highest-level expertise? Second, map the workflow—you provide expertise, facts, and judgment while AI provides structure, polish, and translation. Third, keep what matters—domain compliance, client relationships, professional judgment. Fourth, scale what doesn’t—writing, formatting, translating expertise into deliverables.
If you can explain your work to a smart intern or new hire, you can explain it to AI.
The Economic Reality
Let’s be direct about what this means financially.
If you bill hourly, you’re billing for expertise, not documentation time. AI doesn’t change what you charge—it changes how many billable hours you can actually deliver.
If you work on commission or contingency, volume matters. More proposals, more pitches, more listings means more closings. AI removes the constraint on volume without sacrificing quality.
If you run a service business, capacity is the limiting factor. You can only serve X clients before quality drops. AI raises that ceiling by removing the documentation bottleneck.
If you’re in a marketplace business, speed is competitive advantage. First to market wins. AI makes you faster.
The expertise is the valuable thing. The documentation was the constraint. AI removed it.
What you do with that is up to you. Some people will handle 10x more clients. Others will serve the same number of clients but deliver higher quality and leave on time. Both are winning moves.
The competitive window is finite. Right now, most professionals in your field aren’t doing this systematically. They’re still treating AI like a curiosity rather than a capability multiplier. That won’t last.
Twelve months from now, clients will expect this level of responsiveness and polish as baseline. Early adopters will have established the workflow, built the prompt library, and integrated AI into their practice. Late adopters will be playing catch-up.
The breakthrough isn’t the technology. It’s the realization that your expertise can finally scale without hiring, without cutting corners, without burning out.
Start with one task. Use it for two weeks. Then decide if you want to scale it.
Your expertise is worth more than your typing time. Now prove it.
Wait, This Works for Engineers Too?
Here’s the irony: while everyone’s focused on AI writing code, senior engineers have a different bottleneck entirely. The gap isn’t technical knowledge—it’s translating that knowledge into communication.
For a senior engineer or architect, the bottleneck is RFC documents, design docs, and architecture decision records. The unlock is turning technical decisions into persuasive documentation. What you own: architecture choices, tradeoff analysis, technical judgment, system design. What AI handles: structuring the argument, translating for mixed audiences of engineering, product, and executives.
For a tech lead or engineering manager, the bottleneck is performance reviews, 1:1 notes, hiring documentation, and team updates. The unlock is turning observations into structured feedback in minutes. What you own: people judgment, career development, team dynamics, hiring decisions. What AI handles: turning bullet points into developmental feedback and calibration documents.
For DevRel or solutions engineers, the bottleneck is tutorials, documentation, and customer-facing explanations. The unlock is turning technical knowledge into accessible learning materials at scale. What you own: deep technical expertise and knowing what actually matters. What AI handles: multiple explanation levels, different learning formats, and example variations.
For CTOs or VPs of Engineering, the bottleneck is board updates, investor communications, and strategy documents. The unlock is turning technical reality into executive narrative. What you own: strategic vision, technical risk assessment, roadmap decisions. What AI handles: translating engineering constraints into business language.
The pattern holds: your deep expertise is the valuable part. The documentation and communication layer is the bottleneck. AI removes it.
You already knew AI could help junior developers write code. But it can help you scale the parts of your job that aren’t coding—which, at senior levels, is most of it. A principal engineer spending 8 hours writing a design doc that 3 people will skim isn’t a good use of a $300K/year brain. The design thinking is the valuable part. The LaTeX formatting and prose polish isn’t.
Senior technical roles are communication-heavy. AI just made that communication 10x faster.
I make this Substack thanks to readers like you! Learn about all my Substack tiers here
Subscribed
