Tech Industry
6h
1014
How bad is it in meta?
World Conflicts
Yesterday
263
Since people still can’t wrap their minds around the fact the Gaza death toll hasn’t changed
India
4h
230
India elections - Who is a better leader?
Tech Industry
Yesterday
3562
Job market is brutal for SWEs 🥲
Tech Industry
Yesterday
1689
Meta E6 offer.
UPDATE: the replies on this thread highlight exactly why most people never do anything exceptional. If your default reply is “you can’t do this” “you’re not that good” “I wouldn’t work with you” that’s the exact sort of attitude that reflects on who you are as a person. I’ve heard shit like that my whole life and have a long long track record of wiping the smugness off such faces my whole life, while those same people go back to wallow in their misery. Ask yourself why default to negativity when you could instead choose a better path that leads to more personal fulfillment and happiness. Also, I’ve added a QA section at the bottom. From my personal anecdotal experience over many years, companies, management styles, what appears obvious to me is that there is a widespread misunderstanding how the power law applies to engineering skill and how and when teams and the sizes and personalities of their of compositions are formed. Why teams are overrated? More individuals means more bottlenecks in decision making and in certain contexts eventually crosses a point where a sole exceptional caliber IC would vastly outperform them including an entire department in some cases. Engineering level guidelines are built to capture average engineer levels, they should really capture sr engineer vs multi-domain expert technologist IC. Here’s some examples from my real world experience that illustrates this. 1. Numerous times there’s an engineering team supporting/building/maintaining something that is garbage and rewriting the app would take much less time and result in at least 10x better value (either perf/cost/etc). Yet people entrench their “turf” leading to stagnation. 2. Having a team of average/good engineers build something highly complex when there’s an IC with extremely high domain knowledge and execution speed who could single-handedly build something better, with less time, less bugs, and much higher performance and comes with a uniform coding style. 3. FTX uses like 30 engineers to do what Coinbase does with 500+, and SBF agrees with my thesis (see his twitter rant on it haha) Where teams are justified in many cases, is such as when you don’t have that IC superstar or the complexity of it requires multi-domain engineering experience or advanced R&D topics, or the level of engineering talent is low to implement with lots of toil and lots of rote tasks where failure isn’t a big deal. Not to brag, but to add context, I’d consider myself a top 0.01% engineer. Here’s some real life anecdotes. I’d define this level as multi-technology stack with expert level ability and world class system design skills and exceptionally fast coding execution speed and extraordinarily low level of bugs or performance gaps. I’ve built complex sql models, schemas, queries for modules with complete test suites in a week that similar work took a team months to do, and I had virtually zero bugs, the other group’s was always falling apart. I built an app over week and setup the infra for it (k8s, flux, go, sqlDB, cache) to run it production quality level, something our org couldn’t ever get built right and had been middling with it for almost a year. I solely built and architected and executed our orgs most complex product lines, largest by revenue, profit, and operational load. The amount of bugs/patches needed for this work was extremely negligible and many critical components have never had any bugs, in-spite of how terrible our orgs infra and tooling was and still is. I also revamped alerts and optimized the resource allocation to make the product industry leading in perf SLAs. I quit working for other companies to start my own tech company since I was so tired of dealing with how inefficient supposedly elite engineering “orgs” are with “top talent” and having to manage up on many staff+ engs and first line and sr managers or explaining how/why poor engineering designs and management decisions and practices cripple org efficiency and currently have lots of seed/angel commitments and Tier1 VC interest). $10m+ likely seed for when I’m in active fundraising mode. I’m building my niche in cutting edge cloud computing and infrastructure products. One of its killer features is how it can go from zero to creating fully managed custom micro services that are cloud agnostic kubernetes workloads in a week which take others months to a year to do. tldr; if you’re top 1% percentile relative to big tech engineering peers just start your own thing, or you’ll just get annoyed and quit eventually when your output is 10-100x more value add than your peers who earn similar comps. Q/A How do you quantify skill is a great question. Here’s the metric I use f(n) -> product. Now score these items 1. Speed to implement 2. Bug count and severity 3. How flexible and forward was the design. Would a better one have solved tomorrows problems, would it requiring rewriting your app? 4. How quickly can a sr engineer with no context fully understand what you’ve built and make meaningful contributions. 5. What is the business value (time saved, business won, costs reduced, etc) 6. How much engineering time will it require to operate 7. Cost/Performance/scalability efficiency curves. Plot these metrics 8. How fast can you make changes to this 9. How fast can you identify the root cause to any bugs 10. How well are your tests designed. How easy are they to understand and build from. 11. How easy is it to test your design 12. How resistant is your design. Eg is it very easy to break? How well does it handle fault conditions. 13. Does your design accelerate development speed or hinder it. 14. How much of your app can be run stateless 15. How well isolated and uncoupled is your app to others 16. How well does your app handle versioning and backwards compatibility 17. How well designed are your app metrics 18. How many problems can you solve at once. Eg can you solve broad classes of similar problems with a single design? Additionally engineering mastery includes 1. Accurate schedule planning. I’m extremely good at project management and delivering on time. If you know how to solve something end to end and have experience to foresee reasonable bottlenecks you’ll be able to do what virtually no one else can; eg deliver an accurate engineering timeline. 2. Related to item 1. Ruthlessly cut scope. Understand what’s important and hard about what you’re doing and do that first. Engineering skill hierarchy base-> top 1. Fundamental control eg conditional / loops /fns 2. Using item 1 to solve simple problems in a module 3. Building a standalone module 4. Building a multi-module app 5. Building a system of apps that interact 6. Building a distributed system component 7. Building a distributed system 8. Building complex distributed systems that interact with other outside distributed systems 9. Building complex distributed systems spanning multi cloud, multi region, for item 8 10. Building item 9 on p2p networking components that are Byzantine fault tolerant 11. Inventing new technology such as semi or fully autonomous computer system networks that can self replicate and deprecate and mutate dynamically and can communicate between themselves and coordinate. Eg building an evolution system into computers. What my journey to build my skillset looked like Reading technical books, blogs, and being highly passionate and interested in tech and building like crazy. Not a perfect proxy but I’m usually at 2k+ git commits a year and my Jira ticket completion rate was way higher than anyone I’ve worked with at any company. I can frequently code for 14 hours straight with minimal interruptions or distractions at 100% focus. I’m also an inventor with many trade secrets and open public works inventions. I’m also never satisfied with mediocrity or the status quo and bold and have strong opinions but extremely willing to change them and not afraid to admit being wrong. Also I’ve previously founded companies where I was doing a hell of a lot of coding sometimes reaching 100 hours per week. #tech #engineer
You sound funny and delusional. We will see how you sell this micro service product without team by yourself being 1%. Even Linus didn’t made and Linux alone. Maybe you are 10000000x linus. Although I think it’s more 1/ of that number
I’m not planning to operate it by myself. I think a good head count in engineering would be about 10. The product already has sales pipelines lol. I also have co-founders with the complimentary skill sets. I’m expecting headcount for biz/ops/sales to reach 10-20 over the next year or two
The reason I’m so efficient is because I’m extremely good at abstractions and templating coding patterns. That’s why I’m taking things that are truly awful for enterprise customers like helm charts and offering a 10x better alternative
This is one of those posts I read, and immediately pray I never have to work with the author.
Rest assured no prayers are needed. You aren’t the type of person I’d want to work with. The work environment culture will be largely IC empowered, with uncapped bonus levels eg if a staff engineer proposes and executes an innovative product that leads to new business lines. An uncapped bonus meaning that could be 10m+ for example relative to like a 50-100m arr business line
Also it’s a technology first company, where all the best ideas and proposals drive the company. Regardless of level or function. An intern proposing something better than a sr staff engineer will have their proposal implemented. There’s no punishment for that, evaluations are based on merit purely and in an additive way and all successful promo packets will be internally public and everything related to evaluation to be internally public and open to forum for changing
Is your TC in top 1 percentile of engineers just curious
I’d say so especially now with investor checks. CB pay was for me (or would have been without the stock collapse lol)
It depends on WHAT is being built. Homogenous, high performing teams run circles around everyone else for KNOWN solutions. What FTX did is amazing. But remember they were implementing known patterns/systems foe trading and risk.
lots of founders in silicon valley have these type of attributes. there’s a reason why founders can build incredible companies and run circles around big tech in many instances
How many YOE do you have OP? Curious to know.
~7 years. I’ve probably spent at least 60-70 hours on average doing swe tasks (coding, reading books) during those years. I read around 20 new technical books a year not including re-reads, frequently read many tech blogs, have a track record for invention (trade secrets, and public works)
7 years experience, yet still so much to learn.... I can understand your frustration actually, and I agree that if you are truly exceptional becoming a founder or founding engineer is the way to go. But I also doubt that you are as good as you think you are, or that your impact is as big as you think. Although don't get me wrong, I don't doubt that you are very good (if arrogant). But so are many others.... But you are lacking in experience, wisdom, humility and judgement, and still have a huge amount to learn, judging from this post.
Would love to get in touch.
Send me a DM with a bit about yourself
Done
Why does this resonate so much with me! Especially at a place like Meta where it’s more about showing rather than actually doing
Send me a DM? Probably won’t start hiring till closer to Q1 but maybe you’d like the work and culture im building
If you are as smart as you claim why would you recruit on blind. This is a community of coasters and the non passionate.
I think the problem is that someone has to teach the juniors somewhere. Very often teams simple makedo enough work for the juniors to do something to learn, and then management also wants to justify their own growth and headcount. So you end up going in circles around crappy products that are broken most of the time.
I think that process is broken and am trying a new model for this. Jr engineers will be treated more like apprentices paired with the right staff lead/fit like how renaissance master painters under them who will teach them how to do things right and how to understand the important from the not
The problem is also when the juniors get assigned to other juniors and noone ever brings the good practices. Then the whole team has horrible code quality, yet people on the team don't even have the capacity to recognize such fact. So, any sort of a company-wide questionnaire, gets "great" for "how's the code quality on your team", with only a few outliers from the new folks that just joined.
That’s cos e bottom 99% are Arts major LC monkeys who don’t have any clue what they are here for. That’s what we have major layoff for sw companies since 1990s