My Personal Career LLM with no AI

Introduction

Everyone's talking about LLMs these days. ChatGPT this, Claude that, "AI will replace us all" whatever.

But I've got my own LLM that's been serving me well for years: Leave List Management.

Ever wake up one random Tuesday and suddenly realize your job is complete madness? Yesterday everything was fine, you were cruising along, maybe even happy.

Today you're staring at your laptop wondering how the hell you ended up here, doing work that feels pointless, surrounded by dysfunction you can't fix.

It hits like a truck. One day you're a productive member of society, the next you're googling "how to quit your job" at 2 AM while questioning every life choice that led you to this moment.

The worst part? You're completely unprepared. No updated resume, no network, no savings buffer or interview prep, no clue what you even want to do next. You're stuck between staying miserable or making a desperate exit that could tank your career and finances for months.

This sudden realization forces terrible decisions. You either

suffer in silence until you burn out completely
or panic and take the first escape route you find, accepting a lateral move or even a step backward just to preserve your sanity.

Both options suck. But there's a better way that I discovered way too late in my career. The solution isn't avoiding burnout entirely, that's unrealistic. It's building systems to see it coming and positioning yourself to make informed decisions before desperation takes over.

This is how I built my personal LLM system and used it to make smarter career moves. My hope is that sharing this approach can help other engineers stay ahead of burnout instead of letting it force their hand.

The Leave List Management (LLM)

It took me a while to realize that career burnout isn't as sudden as it feels. The signs are there, accumulating quietly in the background like technical debt. The problem is we don't have a system for tracking them.

For me, burnout doesn't manifest as overwork or crazy hours. Instead, it shows up in more subjective indicators: declining motivation, questioning purpose/progression, non relatable impact etc...

These are harder to measure than hours logged, but they're often more predictive of when I need to make a change.

Every time I start a new job, I create a "Leave List" if I'm feeling less clever that day. It starts completely empty representing the optimism/motivation of a fresh start.

How It Works

As I encounter things that bother me, concern me, or just feel wrong, I document them. Not dramatically or emotionally, just cold, factual observations, like a bug tracker for your job satisfaction.

My goal isn't to build a case for quitting/not quitting or building arguments. It helps me

Track patterns before they become problems
Identify what I can fix versus what I need to accept if you I to stay
Make data backed career decisions instead of emotional ones
Stay prepared for opportunities without feeling desperate

If I can solve issues faster than they accumulate, I know I'm making progress and consider myself "happy". If the list grows faster than I can address items, I know it's time to either escalate my efforts in certain areas and/or start planning an exit.

What Goes on the List?

Everything. Literally. If something gives me a pause, write it down. I was surprised how often things that feel huge in your head shrink when you document them. Ironically, patterns I didn't notice become obvious once they're on paper.

Here's few generalized examples of what actually ended up on my lists over the years:

Technical Frustrations

Senior engineer insists on storing user passwords in plain text because "it's faster"
We deploy to production by SSHing into servers and pulling from main
Code reviews are just rubber with 20 nits without focus on the core logic
Our test suite takes 45 minutes to run and fails randomly 30% of the time and has failed to spot the last few regressions we had
Architecture meetings where we choose technologies based on what sounds cool/trendy, reinventing the wheel in cases we don't need to

Cultural Red Flags

Engineering estimates are treated as commitments, sales/commercial promises are "aspirational" at best
When things break, higher ups asks who's to blame instead of how to fix it
Technical decisions get overruled by whoever is the loudest or has higher reach
We practice 'agile' but haven't shipped anything in 4 months because of 'process'

Organizational Madness

Weekly 90-minute standup meetings where we discuss last week's standup meetings
Three-layer approval process to change a button color, but no oversight on database schema changes
Priorities flip every sprint based on random commercial demos and lack of product vision

Personal Growth Dead Ends

Asked about my promotion trajectory three times this year, got three different conflicting answers
Built a tool that saved the team ~4 hours/week, was told to "focus on assigned tickets instead"
Innovation time is scheduled but indirectly gets filled with 'urgent' bug fixes

...

Making "The List" work

The real power of LLM isn't in the doc itself, it's in what you do with it and how you action on it.

Regular Grooming (Preserving your Backlog/Sanity)

I review my list monthly on a simple calendar reminder with few contingency slots.

Some items will resolve themselves. Others will reveal patterns. A few will grow into deal breakers and start to be repetitive.

I move resolved items to a "Fixed" section. This becomes my personal brag document and shows I'm making an impact (if I am), and make e accountable if I don't.

Influence Analysis

For each item, I question: Can I fix this? influence it? Or is it completely outside my control?

Focus your energy on what you can change. Accept what you can't. And be strategic about what you can influence.

This helps preserve your sanity and focus your energy where it matters most for your happiness/progression.

You'll learn to pick your battles wisely, avoiding conflicts over issues that fall outside your list. I've noticed for me, this approach significantly changes how I reviewed code as I was more likely to accept diffs that work, focusing on core logic rather than nitpicking stylistic choices that could be done differently.

The 2 Columns Rule

I keep two columns:

"I Can Fix This": Items within my direct control or strong influence
"This Is Bigger Than Me": Systemic issues requiring organizational change, the conspiracy theory part!

If column two grows much faster than column one, that's data. Important data.

Exit Criteria

Set clear thresholds. For me:

If the list hits a lot of unresolved items, that number was ~10 in my past experiences
If "This Is Bigger Than Me" has more than 5 items that I know won't move
If I haven't moved anything to "Fixed" in 3 months
If the same types of issues keep recurring despite fixes (moving back and forth between the Fixed section)

Then it's time to start updating the resume.

The Practical Benefits

This system has saved me from bad decisions multiple times:

• Prevents panic exits: Data instead of emotions driving decisions, articulate exactly why I'm leaving and what did I try before going that path helping with the hard discussions • Enables strategic planning: See patterns before they become crises, this can help anticipate problems as well given the commonality of problems we face cross companies • Improve in my role: Many items are "fixable", prioritize battles and frame problems systematically helps me grow as an individual and uplift the team and people I work with • Makes me interview aware: Knowing more what you do want/like/cherish and what you don't is a great funnel for those crowded interviews that will come when need be

LLM revealed for me uncomfortable truths, either:

• The company is actually broken: List growing faster than company evolution means no individual effort will fix it (including mine), especially if most of those are on the second column • I was part of the problem: Realizing my contributions to dysfunction is valuable data I was able to act on

Real World Example

At an old company, my LLM revealed a pattern: every technical decision was made by committee, with the least experienced voices often winning because they were loudest. Architecture discussions became political battles instead of engineering problems.

I tried adjusting it within my control, documenting better practices, proposed clearer decision making processes, rallying people with brainstorming sessions, clear examples & retrospectives on existing systems, even got buy in from senior leadership acknowledging the problem and my efforts.

But at the end, nothing changed.

Worse, the list kept growing with other problems. 6 months later, when a recruiter reached out with an opportunity at a company with strong technical leadership, I was ready. I had specific examples of what I was looking for and what I wanted to avoid.

I didn't leave in anger or desperation and was clear on what I had and why I reached that point.

Conclusion

Career satisfaction isn't about finding the perfect job but more about making informed decisions with the information you have.

The Leave List Management system gives you that information before you need it. It turns the inevitable frustrations of any job into actionable data instead of mounting resentment.

Sometimes you'll use the list to fix problems and fall back in love with your current role. Sometimes it will confirm that it's time to move on. Both outcomes are extremely valuable for your career.

The goal isn't to become a job hopper or a pessimist. It's to stay conscious about your career instead of sleepwalking through it until burnout forces your hand.

Start your LLM today. Open a document, give it a name, and leave it empty. Then wait. The list will populate itself trust me.

And when it does, you'll have data instead of panic. That's worth its weight in equity.

While this approach has proven effective for me and a handful of colleagues, I'd love to expand my understanding with a bigger sample. Feel free to check a template and do let me know

What patterns have you noticed in your own career dissatisfaction?
How do you decide when to fight for change versus when to find a new battlefield?