Microsoft

♦ Led a team of 12 in the architecture, design, and development to deliver a new internal service for all production clouds at Microsoft. Drove the collection of requirements from 4 separate customers through multiple interviews and meetings per customer to understand their use case for VMs used for troubleshooting. We designed a centrally managed and scalable compute solution that enables engineers to simply click a button to login to a fresh and secure virtual machine to test their service or reproduce a customer issue that is automatically deleted when no longer needed. This service replaced 100% of insecure debugging virtual machines from 2 cloud environments and contributed to automated front-end testing for 2 other product teams.
♦ Championed the augmentation, development, and coordination with cross-organizational teams to deliver the internal DevOps platform used to automate the deployment of 100+ services into special cloud environments saving more than 200 engineering hours per month over the existing process.
♦ Reduced network isolation deployment times by 40% by and eliminated the need for two users to initiate concurrent deployments while increasing the completion rate of deployments by optimizing the code running the deployments.

Site Reliability Engineer II

US Navy

♦ Led the design, development, team of 5, and delivery of an application and database solution to solve inconsistent human resource response times and reduce labor-intensive operations.
♦ Communicated with 15+ users to draft requirements and iteratively deliver functionality based on their feedback.
♦ Designed back-end database tables to include fields, data types, and relationships to implement 5+ functionalities in application.
♦ Troubleshot over 700 customer inquiries on application functionality and desired features resulting in 20+ major bug fixes and minimal data loss.
♦ Researched objects and syntax required to deliver automated filling of paperwork cutting 3 weeks of work down to 5 minutes semi-annually.

Application Development Lead

♦ Led a team of analysts dedicating 3,500 hours, analyzing 250,000 results from 200+ queries; resulted in valuable intelligence from 28 previously undiscovered documents, 20% of which yielded 500+ new selectors, increasing the number of known selectors for the team by 25%.
♦ Executed queries on targeted DNI and DNR selectors to satisfy intelligence needs.
♦ Performed tasking, maintenance, and compliance for all targets.
♦ Performed contact chaining to identify the center of gravity in organizations.
♦ Identified new electors for tasked and developing targets

Target Digital Network Analyst

♦ Provided timely SIGINT reporting to the IC through 1,000+ hours of analysis, 150 hours of quality control, and 80 hours drafting 5+ reports. Efforts increased mission production by 200% and established the standard for reporting in the 502 CMT in support of Pacific Commander's objectives.
♦ Drafted and edited initial reports based on quality assurance findings.
♦ Performed both initial and final quality assurance on reports drafted within 502 CMT before released tot he Intelligence Community.
♦ Provided reporting information in response to Intelligence Community (IC) requirements and in accordance with missions of Signals Intelligence (SIGINT), Information Assurance, and Computer Network Operations.

Target Analyst Reporter

♦ Operated signals collection equipment, systems, and antennas for analysis in identifying essential elements of information.
♦ Maintained cryptologic databases to gather additional target information and operating characteristics while protecting sensitive methods and sources.
♦ Supervised signals collection, analysis, reporting, SI communications and assigned personnel.
♦ Allocated Signals Intelligence collection resources and adjusted tasking as required.
♦ Complied with classified information storage, transportation, and protection requirements.

Cryptologic Technician

♦ Located and detected new and sustained targets of interest through search and survey.
♦ Executed Radio Frequency (RF) scans to identify relocated or Off The Air (OTA) targets of interest.
♦ Processed high priority sustained collect signals including both Digital Network Intelligence (DNI) and Dialed Number Recognition (DNR) across multiple field sites.
♦ Identified, troubleshot, and mitigated collection and processing issues associated with sustained mission across two major assets.

Signal Recovery Team (SRT) Signals Analyst Lead - National Security Agency (NSA)

♦ Performed the fix actions necessary, including restarts, configurations, and maintenance.
♦ Performed acquisition, analysis and reporting of state-of-the-art communications signals at the intermediate level.
♦ Provided indications and warnings to national level policy makers through data collection processing and forwarding to the Office of Primary Interest.
♦ Trained operators in the use of multiple follow on processors, signals databases, and other tools needed to complete relocated missions.

Signals Collection and Analysis Operator - National Security Agency (NSA)

♦ Built Infra-as-Code to dynamically generate configuration files for back-end Skype/Teams call routing deployments for more than 5 environments.
♦ Automated feature flag detection resulting in identification and creation of over 1,000 bugs saving more than a month over the current process in the delivery of features for 3 special cloud environments for Microsoft Teams application.

Site Reliability Engineer

Northwestern Mutual

♦ Used financial vehicles to help clients achieve their personal financial goals.
♦ Licensed to sell Life and Health Insurance in the State of Ohio.

Financial Representative Intern

Alex Thielman

Fractional CTO • Cofounder of Tech Leaders • Dad👨‍👧Drummer 🥁MTBer 🚲

In 2021, after exiting my last startup to a public company (founding engineer at Digit.co), I “retired” from Silicon Valley with a plan to never work a traditional job again (but no clue how to make that happen).

Since then, I’ve worked fractionally with 32 different start-ups and scale-ups all over the world the last 5 years as a strategic consultant, trusted advisor, hands-on builder and leader, and everything in between. 

After seeing what’s possible for our teams with the right AI tools and workflows to move faster and deliver higher quality work without wasting time on AI “shiny objects”…

We’ve perfected the playbooks and process for helping individuals and organizations with real AI transformation. 

We can help you/your team:
- Get up to speed with AI and stay ahead of the latest tools, workflows, and frameworks (without the hype)
- Adopt AI tools and workflows for actual ROI (without the fear, uncertainty, and doubt)
- Build and deploy production AI systems that deliver high-quality outputs (even if you’re non-technical)

And we’re happy to share what works (and what doesn’t) with you.

Plus, it’s just more fun to learn together with likeminded folks on this new AI frontier.

Whenever you’re ready:

1. Grab a free copy of our Fortune 100 AI Agent Prompt Library

It’s the complete A-Z roadmap for selling your expertise without burning out. Get it here: https://www.technical-leaders.com/library 

2. Book a Call with Me

If you want some 1:1 help we can jump on Zoom for a quick call to discuss your approach to AI. Book here: https://calendly.com/tech-leaders/ai-strategy-session 

3. Work With Us

I’ll always give away WHAT to do and WHY it works, but if you want our help executing on the HOW without all the trial-and-error… we can help you here: https://www.technical-leaders.com/ai

P.S. Send me a direct message, I love hearing about what you’re up to with AI

Todd Larsen

Companies are stuck on ChatGPT like offices in 1995 were addicted to their all-in-one printer-scanner-fax machines.

Sure, it does everything...
Just not particularly well.

Here is a prime example:
→ My client burned through six months trying to make ChatGPT work for their technical documentation workflow before calling me
→ Their problem wasn't the AI. They were using consumer software (ChatGPT) for professional work

This is like trying to run a construction site with tools from Home Depot's bargain bin

→ ChatGPT writes decent emails
→ Claude structures complex analysis
→ Perplexity handles research depth
→ Gemini processes multimodal data

Each tool built for specific professional outcomes. Not general adequacy.

The shift happened when we mapped their actual workflow to purpose-built AI tools

Documentation time dropped from 3 hours to 45 minutes per project

Quality scores jumped 40%

ChatGPT got everyone started.
But professional work demands professional tools.

What's your most time-consuming workflow right now? Drop it below, I'll map the exact AI stack that cuts it in half.

AI-First CTO • Helping Human Intelligence Lead Artificial Intelligence • Dad • Deadhead • Drummer

A client just doubled their consulting rate after we stopped focusing on their implementation skills entirely.

Instead, we mapped their value to the questions nobody else was asking.

Sure, AI can write a payment gateway in minutes.
But can it tell you why your checkout flow loses 47% of customers at step three?

AI can optimize your database queries.
But can it spot that your entire data model breaks when you hit enterprise scale?

They translate between what's possible and what's profitable.

Start by asking better questions:
→ What happens to your margins when you scale this?
→ Which technical debt will kill your next funding round?
→ How does this architecture decision affect your acquisition value?

The market doesn't need more implementers.

The market desperately needs engineers who think in systems, speak in outcomes, and build for leverage.

While everyone else competes on how fast they can code, the real players compete on how clearly they can see.

how to design AI workflows for the work you don’t want to do

I’m looking for 6-7 people
Who are looking to deliver an AI-first project in the next 30-60 days
Without wasting time or getting too technical

So if you or someone you know:
- Want to use the best AI tools and workflows (beyond ChatGPT/Cursor/REPLit)
- Need to deliver on deadlines faster without getting called out for “AI slop”
- Have a new idea or current project you’re focused on delivering to customers/stakeholders in the new year

Like or comment on this post and I’ll send you the details

I used to think the secret to AI success was finding the right tools.

Spent months testing ChatGPT, Claude, Cursor.
Built a ton of half-finished workflows that never shipped.
Watched my team get excited and then slowly lose interest.

The problem wasn't the tools.
It was discovering a new tool and jumping straight to "How can I use this?"

If you skip strategy, you'll end up with shiny experiments that deliver zero ROI.

MIT recently found that 95% of corporate GenAI projects fail.
But the problem wasn't the technology.
It was that teams never answered WHY it mattered before they started building.

The engineers and consultants I work with now start differently:

→ Why does this problem cost us time or money?
→ What does success actually look like?
→ What risks exist if this fails?
→ Only then: How do we build it?

One client used this approach to automate their proposal process.
They didn't start with "what AI tool should we use?"
They documented their current workflow first, identified the human-critical steps, then mapped AI to the repetitive 80%.

Result: 6 hours per proposal down to 45 minutes.
Quality stayed high because humans stayed in the loop.

The question isn't whether AI will transform your work.
It's whether you'll build systems that actually stick, or keep chasing tools that fizzle out.

A lot of senior leaders wonder why AI adoption fails to make an impact

But their OWN personal AI usage? Often zero.

You can't navigate by hyped-up headlines while expecting your teams to deliver miracles.

The companies winning with AI don't have executives who just talk about it

They have executives who use it themselves. Daily, personally, and publicly with their teams

Your team will never go further with AI than you're willing to go yourself

Your AI experiments keep failing because you're solving the wrong problem.

Here's what's actually happening:

→ You see a new tool, get excited, and immediately think "I could use this for X."
→ You build something quick, test it once, then it sits unused.
→ Three months later, you've got a graveyard of half-built workflows and zero ROI.

Sounds familiar?
A lot of technical leaders I work with skip the foundation and jump straight to implementation.

So I've created a framework for building workflows you'll actually use:

1. Document your current process first
Write down every step of one repeatable workflow you do manually.
Don't optimize yet, just capture what already exists.
This alone is valuable because most expertise lives only in your head.

2. Apply the 10-80-10 rule
A human provides the first 10% (input and direction).
AI handles the middle 80% (the repetitive heavy lifting).
A human closes the final 10% (review, edit, verify).
This keeps quality high and humans in the loop.

3. Start with ONE workflow
Pick something you do weekly that takes 2+ hours.
Don't feel like you need to use every new tool.
Build one system that works, then expand.

The teams I work with who follow this approach build workflows that compound: faster proposals, better documentation, scalable systems.

What's one workflow you could document this week? Get started with the first step, and see where it takes you.

taplio

The Battle

Alex Thielman

Todd Larsen

Followers

Posts

Engagements

Likes

Performance ScoreTodd Larsen's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Alex Thielman

Todd Larsen

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Todd Larsen's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreTodd Larsen's LinkedIn Performance