Aaron Linder MSc

🚀 10 Signs You're a 10x Engineer (and don't even know it!) 🧑‍💻 1. You've googled "how to exit vim" more times than you'd like to admit. 2. Your commit messages are works of art. Sometimes they even rhyme. 3. You've fixed a critical bug by adding a single semicolon. 4. Your code comments are a perfect blend of humor and despair. 5. You've mastered the art of looking busy while waiting for your code to compile. 6. Your rubber duck has seen things no duck should ever see. 7. You can explain technical concepts using only memes. 8. Your coffee intake directly correlates with your git commits. 9. You've had heated debates about tabs vs. spaces. 10. You've written "it works on my machine" so many times, you're considering getting it tattooed. Bonus: You're reading this and nodding vigorously! 😄

82

16 Comments

Are colleges/universities failing Computer Science grads in 2024? The vast majority of CS grads end up working as software engineers. But many CS grads are unable to answer basic programming questions during an interview. (We talked about OOP last week). Why should CS grads have to invest in a bootcamp or additional hands-on training to land a job after a pricy college degree? We provide these hands-on skills at no cost at HTD Talent, but we can only help so many. Has this been your experience? What do you think is behind it? Let me know in the comments.

42

19 Comments

The opener of "Beyond Cracking the Coding Interview" will be a section on why technical interviews (and job searches) suck. That may seem odd, given that the book is ENTIRELY about how to pass technical interviews and manage your job search. But I'm convinced the reason many people don't do well at these things is because of fear and discomfort. You can't effectively prep until you make peace with the suck. And making peace with the suck requires owning that the process is painful and bad and flawed. Below are some reasons we have already for why technical interviews are job searches suck. What are some others? ------------- Job searches are not deterministic, and neither are interview outcomes. For job searches, there’s often no correspondence between how much effort you put in and what result you get out. For technical interviews, there’s little predictability in how the same person will perform from interview to interview (no wonder many people are frustrated with the format). No feedback loop. When you apply online and don't get a response, you can't tell if you weren’t a fit or if no one even saw your application. Your insecurities can convince you that not only did a human look, but they quickly sized you up, saw right through you, and lasered in on every single flaw to conclude (correctly, in your mind) that you’re totally unfit for the job. When you interview, whether you pass or get rejected, you often don’t know why, which makes it difficult to know how to prepare the next time. The content of your resume is often eclipsed by the brands in it. If you don’t have brand-name companies or schools, it’s much harder to get noticed. Recruiters are notoriously bad at making value judgments based on resumes. Despite these shortcomings, resumes are still the gold standard, and that means candidates from non-traditional backgrounds enter the game with a significant disadvantage. To get in the door, you very likely have to know someone. Surprisingly, this is even true for candidates who look good on paper; recruiters often ignore online applications because the signal to noise ratio is so poor. But it’s especially true if you don’t look good on paper. Technical interviews are notoriously flawed and not representative of the actual engineering work you do every day. This one is especially rough, and it bears out in our data. Senior engineers often do worse than juniors in their first few interviews because junior engineers have more recently completed an algorithms class or have done extensive interview prep. Senior engineers have been in the trenches, often focusing on building applications; there are very few engineering roles where you’re doing the types of academic problems that you get in interviews day in and day out.

34

4 Comments

Your new hire might be on their way out sooner than you think: Wondering why? According to SHRM, 20% of new hires leave within the first 45 days. But why? Bonus: Sharing Editable Resume template to get dream jobs Meta Resume Template for Software Engineers Get it here: https://lnkd.in/gU6QZsBu 🍏 Apple Resume Template for Software Engineers Get it here: https://lnkd.in/gW-8ab9z 📦 Amazon Resume Template for Software Engineers Get it here: https://lnkd.in/gP-YZyfD 🍿 Netflix Resume Template for Software Engineers Get it here: https://lnkd.in/g6FWDNZF There are several reasons, including: ➜ Inadequate training ➜ Lack of feedback ➜ Poor integration into the team ➜ Job not meeting expectations You can change this trend. Employers with a well-structured onboarding program see a 69% higher retention rate over three years. (Source: LinkedIn) The secret? Support from day one. Here's a roadmap to take care of your new hires: Days 1-30: ➜ Meet the team. ➜ Complete necessary training. ➜ Set initial goals. ➜ Learn company values. Days 31-60: ➜ Grasp key performance indicators (KPIs). ➜ Discuss performance progress. ➜ Expand your internal network. ➜ Start small projects. Days 61-90: ➜ Propose solutions to problems. ➜ Continue skill development. ➜ Stay organized to meet deadlines. ➜ Keep building trust with the team. Days 91-120: ➜ Review overall performance. ➜ Optimize workflows. ➜ Seek continuous feedback. ➜ Plan for long-term objectives. Don’t just onboard—educate. Don’t just onboard—connect. Don’t just onboard—empower. Credit:- Justin Mecham and Want a PDF of his top infographics? 👇 ▶️ Go Here: https://lnkd.in/g2xbnwhp Please repost to help others! ♻️

77

30 Comments

Scrutinizing Mercor's Hiring Proposal: Do the Numbers Add Up? Mercor, a San Francisco-based HR technology company founded in 2022, has recently advertised positions offering $210 per hour for 100 senior engineers. While this appears attractive, a closer examination raises concerns about its financial feasibility. Financial Breakdown: • Annual Compensation per Engineer: • 40-hour workweek: $210/hr × 40 hrs/week × 52 weeks/year = $436,800 • 80-hour workweek: $210/hr × 80 hrs/week × 52 weeks/year = $873,600 • Total Annual Payroll for 100 Engineers: • 40-hour weeks: $436,800 × 100 = $43.68 million • 80-hour weeks: $873,600 × 100 = $87.36 million Mercor's Financial Standing: • Estimated Annual Revenue: Approximately $12.9 million • Total Funding Raised: Around $33 million • Employee Count: Approximately 93 employees Analysis: The proposed annual payroll for 100 senior engineers, even at the lower estimate of $43.68 million, significantly exceeds Mercor's current annual revenue of $12.9 million. This disparity suggests potential challenges in sustaining such salary commitments without substantial increases in revenue or additional funding. Considerations: • Revenue Alignment: Is Mercor's revenue model structured to support this level of payroll expenditure? • Operational Sustainability: Can the company maintain financial health while committing to these salaries? • Strategic Planning: What growth strategies are in place to justify and support this scale of compensation? While competitive salaries are essential for attracting top talent, it's crucial for both prospective employees and stakeholders to assess whether such compensation structures are sustainable within the company's current financial framework. A thorough evaluation of Mercor's financial health and strategic plans is advisable to ensure long-term viability. Note: Financial figures are based on available estimates and may not reflect the most current data.

I'm trying to get ahold of engineers at Gather for their open video engineering roles, but I will not be paying for linkedin premium. I'd rather be a little scamp/thug/gangsta and call them out publicly. Yes I want to fill one of your upper level engineering roles in video streaming. Yes my experience aligns well. Yes I like the product you make. They want somebody with webRTC experience + video call app experience. That's the catch. I don't have that. I do have this however: 1. Multi million dollar impact on video transcodes that lead to the first 1 billion dollars made at Pluto TV. 🕺 2. Have recently built complex lower level video primitive work in Rust, which you can check out here: https://lnkd.in/epPysFv7 3. Have experience working with third party libraries to provide fallback and distributed scale to ad transcodes with volumes leading to revenues in excess of 1b usd. 4. Have a propensity and ability to work and build high volume distributed notifications systems (or work with them, I'm cool with that!) 5. have literally dove into shark infested waters to help maintain code nobody else understands for a client, providing an asynchronous refactor of code for communicating with nixon era mainframes in a dead dialect (so Im pretty sure webrtc will be easy in comparsion, letsbereal) The ball is in your court. Now network, Let's make this lil job-prosal as awkward as possible and like, comment, share, and tag as much as you can so they know who I am by like Tuesday. Let's make a ruckus and be real annoying. Even if they end up not going with me it'd be neat for other people to check out my skills. 🦀 🤝 🍇 #gather #networking #letsgetannoying #linkedinlunatic #Itakethingsintomyownhands

57

35 Comments

Is your interview process your biggest bottleneck when hiring engineers?? Stop sending take home challenges to engineers or you will lose the race! The market is heating up and you don’t want to be the only startup who is still sending coding take home challenges I get it, this is a critical hire and you need to get it right But sending a take home challenge is the wrong approach After working with 100’s of startups, below is the most effective interview process IMO 🎯1st stage: Your recruitment partner or talent function should be conducting this, if they are not seriously consider making a change 🎯2nd stage: Coding fluency and system design interview (2hr) 🎯3rd Stage: Coding challenge- evaluating how the engineers solves problems, ways of working and the questions he/she asks 🎯Final: Meet CTO/Founder- cultural check and answer any questions they may have Depending on availability etc this process should take no longer than 7-10 business days If you need any guidance on this please reach out as happy to help P.s. go easy on me, this was my first video and it was completely off the cuff 😅

14

9 Comments

There are many "consultancies" out there who ask for the last 4 of your SSN, even before you've interviewed with them. Please beware of them, especially new grads/or folks desperate in this market. The only good time to share it is when the offer is on the table, and a background check is to be done.

29

1 Comment

It took several attempts at modifying rpicam app to get this working and I'd say its 95% working.  What is it? I wired in the Hailo image detection post processor to a new RecordingManager class, and then (this was the hard part) I got libav to write mp4 files for the parts of the stream just when object detection's are happening. By default, rpicam only supports writing to file constantly or not at all. I just tested it out and it finally works --In 15 minutes of me leaving to get a coffee, this is the only two videos it took!   Now the missing 5% of my effort is that these mp4s aren't perfect containers and cannot be opened by all apps, but fmpeg's ffplay is handling them just fine. There are some warnings in the ffplay output for me to debug, but after many failed attempts at adjusting the libav muxer on the fly, with this, I am super stoked. There were a few challenges here: * libav is invoked for the camera device to create a stream muxed to a preview window, the post processor and possibly a file. * libav segmentation, ring buffer and the like are not implemented for Rpi5 * Post processing runs in a separate process from the main event loop of the camera app where we have access to the existing encoder context, so I used a shared /tmp /file to share state from the RecordingManager to the main app to toggle the recording functionality.

18

3 Comments

I’ve reviewed 4,000+ software engineering resumes over the past few years and I see so many engineers failing to land an interview for the same reason: Their resume looks exactly the same as everyone else’s. I don’t mean stylistically. I love a good plain text resume with bullet points. What I don’t love is seeing nearly word-for-word the same “accomplishments” that don’t make you stand out at all. Which one of these is more likely to get a hiring manager’s attention? 1. “Actively participated in software design and architecture discussions, providing expert guidance and insights to enhance project outcomes and technical strategies.” 2. “Led software design and architecture decisions, enabling a 20% faster go-to-market for key projects and a measurable improvement in system reliability and performance.” Anyone can participate in design and architecture discussions, but what makes YOU effective at these discussions? Can you put a measurement to your effectiveness? While not everything in your workday can be measured, hiring managers are looking for high impact. The next time you update your resume, ask yourself: does this bullet point highlight the unique value I bring to the table? If not, it’s time to rewrite it.

200

61 Comments

There are two really hard things in Computer Science. Cache invalidation, and naming things. That's why I cringe every time I see people abuse those names without giving it enough thought. Stand-ups where everyone is sitting? Post-mortems that just cast blame instead of looking for root causes? Sprint 'planning' that's actually just leadership telling everyone what they'll be doing? Every time you misuse a name, you're making things harder on your team. Communication suffers. Trust erodes as cognitive dissonance grows. Collaboration and synchronization become harder. People have put thought into those names, at least use them correctly or use different ones!

4

Follow here or on YouTube for our weekly Foundry & AIP tutorials!

3

Delta Lake adds more performance enhancements and interoperability features with 3.2, available now! This release includes: * Liquid clustering * VACUUM with file inventory * (preview) type widening * (preview) UniForm write support for Hudi

4

This will probably be an unpopular opinion, but object oriented programming is way better than functional programming. Yes, functional programming probably has a better syntax. Yes, it's probably easier to avoid bugs with FP than with OOP. I don't necessarily disagree with any of these arguments. However, software development is not just about coding. OOP is better because it happens to be the dominant paradigm in software development. It implies the following: - There are way more OOP code samples to look at than FP - There are way more open source libraries written for OOP-focused languages than for FP-focused ones - There are way more OOP tutorials than FP, which means that the OOP-focuses languages are easier to learn than FP-focused languages - Since there are much more OOP samples in LLM training data than FP samples, the answers that LLMs will provide will probably be more accurate if you prompt them to come up with OOP code In short, it's not the cleanliness of the language syntax that determines the relative quality of the language. It's how much faster you can solve problems by using this language compared to another language. And this is where FP languages lag behind.

99

207 Comments

And what is this about? No more runtime fee? https://lnkd.in/gZwkrrMD Do I trust Unity3D enough to not change their mind in the future? No. They definitely made a lot of developers and studios gun-shy about using Unity3D. I think there was more noise than there was action, on the part of developers abandoning the engine. They certainly lost a lot of my trust. Would I consider Unity3D for a project in the future? Certainly. But like all things, there is business risk, and that has to be evaluated. Unity3D "won" a lot of hearts and minds because they were small studio friendly, plus the lower barrier to entry for people who don't do traditional game development. But Epic with Unreal Engine has been a rock solid company to hitch your star too, and completely transparent in their licensing. In fact, it's the kind of licensing that is stable enough you can predict the future around it. Unity3D kinda fucked that up, and constant license changes and special rules around different business sectors make risk averse businesses less likely to adopt the tech. Either way, it's good news, but we'll see how the future plays out.

6

2 Comments

Did you see the new 2024 Stack Overflow Developer Survey that just came out? 🚀 Some insights and key findings: - Top Programming Languages: JavaScript, Python, and SQL are among the most used and desired languages. Rust, however, continues to be the most admired with an 83% score. - Databases: SQLite's popularity has grown, surpassing MySQL in developer interest. PostgreSQL remains a top choice with high admiration and desire scores. - Cloud Platforms: Hetzner leads as the most admired cloud platform, with AWS and Google Cloud being the most utilized. - Web Technologies: Svelte and React are highly desired for web development. Svelte, in particular, has seen a rise in interest. - Top Paying Technologies: Erlang developers reported the highest median salaries, followed by Go and Zig. You can see all the results below: https://lnkd.in/gytSxtwu #TechTrends #WebDevelopment #CloudComputing #DataScience

1