Taro Experts

In my humble opinion, no job in the entire world can make you fulfilled if you don't have anything outside of work to look forward to (family, friends, SO, a hobby, etc).

I have a very simple rule for generating fulfillment from work and that's to find the next opportunity to learn something new or do something you've never done before. It could be leading a large cross-functional project. It could be mentoring a junior engineer. It could be working with a tech stack you've never worked with before, etc

Nothing kills fulfillment more than being in an environment where you aren't learning anything at all or learning something so niche and specific it doesn't translate into anything fundamental (ex. very specific frameworks, etc)

First off, congratulations. That is no mean achievement in this economy.
If you're looking to start up, then in all likelihood your first few months (and even perhaps years) will necessitate that you have a great understanding of the business you plan to run. So, in that case, it is a question of just getting your feet wet if and when you want to.

Alternatively, if you prefer to move to a TPM role, here are my two cents.

I think a good TPM has a combination of three offerings -
1. good technical skills (plain-speak - has written code in the past)
2. great people skills (high empathy for engineers, good understanding of business requirements and an even better translation of business needs into the coding world)
3. supreme prescience (being able to have the right pulses to 'know' the future direction of every single initiative)

You probably have the first skill above. If you're still new to the tech world, a few months in a fast-paced unicorn will help you polish your tech skills. Be sure to check this course out - https://www.jointaro.com/course/level-up-your-code-quality-as-a-software-engineer/what-actually-makes-code-good/. It is probably the best lottery ticket out there to learn how to code 'better'.

The second skill is highly experiential. The more time you spend talking to engineers and people with business needs, the better. Much of communication is also about asking questions, collating information, and having a 0->1 understanding of every system you work with.

The third skill is the toughest in my opinion. Two resources come to mind. A - This essay by Paul Graham - https://paulgraham.com/greatwork.html and this remarkable video by @Alex - https://www.jointaro.com/course/maximize-your-productivity-as-a-software-engineer/finding-the-best-things-to-work-on/. Be sure to check out the matrix that he refers to. Doing great work helps you know what is important. And knowing what is important enables you to predict future direction better!

As for timelines, I do think you should do it at a time when your manager thinks you have contributed sufficiently and you have learnt enough of the above three skills to move.

I've tried like almost every single one and heres my "AI stack"

• cursor for coding. I would not use any other IDE simply because of cursor's copilot. its too good. I havent tried cline/windsurf/aider. but cursor works good enough and their agent is getting nicer and nicer. Plus they now have MCP support so it can talk to your database. It's agent works pretty well where it can iteratively figure out what files to read and what changes to make and recursively keep making changes till tests pass

• cursor also has AI in the terminal so i use that a lot

• Perplexity for all questions I have. pplx has access to the internet + gives access to claude/oai/R1/o3. I use this for coding mostly when I need to update code based on latest documentation or might expect that the answer is in a stackoverflow thread

  The caveat with perplexity is that its not good for general purpose coding questions. They kinda nerf it because they reduce the temp so using native claude is almost always better. But I havent ran into too many issues because cursor is good enough for most things for me

• I am a huge huge huge fan of perplexity. It's literally learning on steroids for me

I actually setup a perplexity space for Taro, so basically it only looks at jointaro.com links which is pretty cool

Everything else has a free usage limits that's good enough for one off queries I might have.

Claude is best for coding/writing

o3 mini for hard problems

gemini from ai studio for long context problems (think 30 pages of PDF papers)

grok from xAI is meh

Do you have an onboarding buddy there or another engineer that's also L5 or above? I'd build a relationship with them and leverage them to get a high-level view of the system and all the tooling etc that powers your specific team's features.

It's totally fair to ask to set up time with them for an initial meet & greet + a get a high-level diagram of the entire system along with the plumbing tooling. At the end of the meeting, you can them if they'd be open to future meetings/questions, and you can use that to ask targeted questions later on.

In my repeated experience, this approach is much more efficient than trying to figure it out on your own (and building up anxiety over low output along the way).

I'd also take extra time in evenings and weekends to ramp up on these peripheral tools and systems to get up and running quickly. I am not a proponent of bad WLB, but sometimes it's just the need of the hour, esp. when performance at new job is at stake.

Instead of paying the $50 on fiverr maybe you should pay me instead 🤔🤔🤔🤔🤔🤔🤔🤔🤔

The experience is okay enough where it should get bites, but the resume formatting is not that good.

First priority: if you're paying more than 1k yearly for the masters program and/or spending time regularly on it, drop it immediately. Your career is too late where a master's program will help out your career and it's sucking up time and money. If I were an employer: Stanford or bust. (Assuming you are a US citizen) No other school quite franky justifies the value for the "oh I want to do a Masters to make myself more marketable" route other than Stanford.

I read your work experience and honestly I just glazed over it. It's fairly verbose (in terms of the wording and the number of bullet points).

• Everbeat: number of bullets points is fine and you clearly wrote code for the stuff. But what was the impact? What does the company/your team do to provide value and how did your work further that?
• Amazon: there's a brazilion bullet points with lots of words. My mind pretty much shut down when I read this (code for "your resume is getting tossed"). Have less bullet points with simple, more concise wording. And focus on impact.

Honestly, just run your resume through ChatGPT for free and ask it to make the wording simple & concise and flatten the number of bullet points. That'd get you the most yield of your time and money.