Guy

Most engineers think their CV is about listing experience. It’s not! CV’s real job is to earn you the conversation in under 10 seconds. This is one of the most detailed breakdowns we’ve published on how to write a CV for Web3 technical roles based on 400+ interviews we conducted in 2025. We breakdown: - how recruiters skim CVs after the ATS filter - how to write a professional summary that signals depth immediately - how to turn roles into quantified impact (TVL, % improvements, adoption) - how to present web3 skills so they’re easy to evaluate - how to showcase projects, open-source work, private audits, and hackathon wins properly - how formatting and structure quietly make or break your CV We’ve also included a free CV template (no login, no gating). If you’re applying for technical roles in Web3 (or knows someone who is) this should save weeks of iteration. Credit to @itsme_madara for doing the heavy interview lifting, spotting the patterns, and pulling this guide together.

touch the grass they said..

We are hiring a Web3 Finance Lead. Remote | $70k–$110k - Strong finance/accounting background with experience owning FP&A, reporting, monthly close, tax, payroll, and treasury operations - Hands-on experience managing cash flow, payments, reconciliations, and financial records across both fiat and on-chain accounts - Comfortable working with external accountants, bookkeepers, tax advisors, and payroll platforms like Deel or Rippling - Bonus: crypto/web3 startup experience, stablecoin payroll, multi-entity structures, and US/international tax exposure Apply now or refer a friend! (Link in comments)

@RenanRSouza35 @Jeyffre I've noticed this too

@Jeyffre Have you tried yoga? it completely changed my sleep quality

After extensive experiments on myself (n = 1), I've found the key to extended deep REM sleep: - eliminate coffee completely - walk a minimum 6,000 steps a day To clarify: - No coffee doesn't mean no caffeine. I still drink a lot of tea / yerba mate. Something about coffee specifically seems to mess things up. Honestly sucks because I live on an island named "Java" and it literally has the best coffee in the world. I'll still drink coffee socially, but as an exception, not the norm. - oddly enough, substituting other exercises for walking like lifting weights doesn't do it for me. I installed a gym in my office. Lifting weights weekdays and running a 5k on the weekends didn't solve inconsistent sleeping. Now here's the surprising stuff. The no coffee + extended walking means I still sleep well despite: - I sometimes work until the moment I go to sleep, and I don't use blue light glasses. Sometimes I'm doing pretty intense math right before sleep. Doesn't affect me anymore. - even crazier -- I sometimes go to sleep while working. I'll have deeply technical discussions with an AI while scribbling notes in Notion. I still sleep soundly. - I don't pay attention to the timing of the last meal of the day; that has had little effect on sleep quality for me - I don't take magnesium or other pills before bed. It has a slightly positive effect, but it's a rounding error compared to the two items above - sometimes I play sports up to 9pm or so. Generally considered bad form to exercise at night, but that hasn't ruined sleep for me. Here's other stuff I experimented with: - blue light glasses only have a minor effect on sleep quality, at best - whether the economy / business is doing well or bad has little correlation - avoiding processed foods / eating healthy helps having focus after the meal, but doesn't affect sleep - totally abstaining from alcohol has little effect - turning off screens an hour before bed and using only dead trees has little effect - bed time has a weak correlation. Obviously, going to bed at 2am isn't great, but the difference for me between 10pm or 11:30pm is when I'll naturally wake up later, not how well I'll sleep. Obviously, I have to put in the disclaimer that I'm not a doctor and n = 1. But if quality sleep has felt somewhat random to you, I can recommend seriously increasing your step count and replacing coffee with something else. Both of these remedies are admittedly a little inconvenient. Walking for a long time means I have to figure out how to get work done while on the move, but it's worth it.

Maybe it will become trendy again to optimize your software to use less RAM.

Most Solana test output answers one question: Did the transaction succeed? I'm interested in a different question: What story did the program just tell? A new capability I'm working on lets downstream tools reconstruct execution as a behavioral trace instead of log soup. (Example: @C4ngui's vesting program) 🧵

We are hiring a Senior Data Engineer (Remote). - 7+ years building production data infrastructure, ideally in trading, fintech, or low-latency systems - Strong Go or Rust for data services, with Python for pipelines and analysis - Deep experience with streaming systems, time-series storage, event sourcing, CDC, Kafka, or Redpanda - Strong correctness mindset across ordering, reconciliation, idempotency, exactly-once semantics, and point-in-time consistency - Bonus: CEX/DEX market data, on-chain indexing, margin, PnL, risk data, ClickHouse, DuckDB, Iceberg, Snowflake, or lakehouse tooling Apply now or refer a friend! (Link in comments)

I’ve tried learning Rust many times and got derailed for one reason or another. When I started @RareCodeAI, I finally stuck with it. The focused learning path made the pieces click, and before long I was writing utilities on my own. It’s a great launching pad for anyone looking to get productive with Rust. Thanks @Jeyffre 🙌

This week, we added two new articles on NTT to our @RareSkills_io ZK Book. Four more to go. NTT is a fundamental tool in ZK - and in many other areas - and we hope our detailed tutorial will be useful for anyone learning it. Last month, we also started a Halo2 bootcamp, so material on Halo2/PLONK/Plonkish should start showing up soon. Good things are being cooked. In the meantime, if you want to learn ZK with us, there's a Fundamentals of ZK cohort starting in July. We also have Advanced Solidity and Rust bootcamps scheduled for July.

It may be self-aggrandizing for me to say this... but I think ZK may legitimately be the hardest applied subject in computer science. As the post below notes, just to even start, you need a background in abstract algebra, cryptography, and theoretical CS. Most programmers don't even take courses in these subjects (and likely didn't master them even if they did). Just to read ZK papers, you need a strong background in computational theory (which most CS majors don't). And the annoying thing is that modern ZK relies on computational theory results that are newer than when a lot of the good textbooks were written (e.g. the textbook predates IP = PSPACE or Probabilistically Checkable Proofs, PCP). Just understanding IP = PSPACE or PCP is genuinely hard, and those are just table stakes for some papers. Another blocker: polynomials aren't bad, but in my experience, even programmers good at math have long forgotten that the degree of the product of two polynomials is the sum of their degrees. Forgetting "base level" knowledge like this makes navigation harder. And I don't blame programmers for not knowing this law, because you almost never use it in real life. There's also a bunch of random polynomial identities that most elementary courses omit. Probability and combinatorics aren't heavily used, but they pop up enough times that you'll get jammed if you haven't formally studied them before. Now let's get to understanding FRI (ZK-STARKS). Most engineers don't know the Fast Fourier Transform, and even if they do, they don't have a clear picture of why it works. Not hard to learn, but it's another obstacle. Then we get to coding theory. Coding theory is easy to visualize, but the major open questions in the field were not resolved until recently, so many textbooks have generally not caught up. This is one area where I've found even frontier LLMs to not be helpful at explaining things because they have very little data to go off of. Topping it off, most devs don't even know what coding theory is to begin with. Ooh, we ain't done yet. No ser, that was just the warm-up round. To genuinely understand the paper "Proximity Gaps for Reed–Solomon Codes" by @EliBenSasson et. al. you need to know some algebraic geometry, which relies on commutative algebra, which comes after having mastered elementary abstract algebra. So you're at least three levels deep into a subject that only math majors (maybe) touch. Thankfully, simpler proofs for proximity gaps exist, but those proofs weren't discovered until at least 2022 or 2025, depending on what paper you consider qualifies. I'm just including this to show that the math can get genuinely nasty. There isn't a "finish line" you can cross where things suddenly get easier -- especially since the finish line moves every year. Furthermore, everything of consequence is written in Rust, which honestly isn't that bad, but it still takes a good 1-3 months of grinding to get fluent with Rust, as you can only internalize so much syntax at once. Then, to do ZKVMs, you need to be comfortable with computer architectures and low-level programming. Again, those aren't that bad, but it's something most programmers aren't already fluent in because it's not "necessary" for a lot of jobs. In comparison, machine learning more-or-less stops being difficult if you have your linear algebra, statistics, and calculus down. Then it mostly turns into engineering/experimental problems. I'm speaking from experience with this. I rarely found ML to be hard, and most of the time I found it geometrically intuitive. ZK is hard because it goes wide and goes deep. It draws on many subjects and relies on frontier results from several of them. Now, to keep some perspective, every subject is easy in retrospect -- once you do it enough times to gain unconscious competence. But when your efforts are spread across so many disparate subjects, and you have to go deep into many of them. Therefore, your reps get spread across more areas, and there isn't much data on these topics to feed the LLMs. You can grind Zero Knowledge Proofs for years and not feel like you've really mastered it. ZK 100% not a shortcut to making a good salary. But if you are inspired by a challenge, energized by a battle of wits, and can take 11 KO punches and get back on your feet for round 12... ZK might be for you.

1 - So GLM 5.2 is 700b parameters (ish) 2 - 4x DGX Sparks can supposedly handle up to 700b parameters (give or take) 3 - GLM 5.2 is supposedly in striking distance of the performance of GPT 5.5 and Opus 4.8. In my brief tests, it's really not shabby at all. 4 - So for $20k, you can get near the frontier on your table. 5 - Extrapolate the trend, and you could have mythos/5.5 pro - class models in your dining room for the cost of a cheap car less than five years from now. Even without extrapolation, we're already the near frontier running locally. 6 - Paying real api costs, I could easily blow through $3,000 per month coding and running agents. The machine pays for itself in 6-7 months conservatively. 7 - In 3-5 years, most power users of AI will self-host. 8 - Am I missing something?

Learning Solana through @solanaturbine. The first wall wasn't understanding accounts or CPIs. It was reading the damn logs. Flat transaction logs work great for grep and CI pipelines. Not so much when you're debugging at 1am.

If you want to learn how to write or audit ZK circuits, learning Circom is a great place to start. It is arguably the easiest circuit language to learn, and the RareSkills ZK Book includes an entire module on it, from the basics all the way to building a simple zkVM. That said, Circom is not the most suitable language for zkVMs. Last month, we started our first Halo2 bootcamp, which combines PLONK arithmetization with custom gates and lookup tables. Halo2, together with Plonky3, has become one of the most widely used libraries for zkVM development. We hope to update the ZK Book with these technologies soon. Even so, Circom remains the best language for learning how circuits work, and it is entirely possible to build privacy-preserving applications using only Circom and Groth16. Don't forget to check out the RareSkills ZK Book.

If you're starting to learn ZK or cryptography, getting familiar with a finite field library is important. One that we use extensively in our courses is the galois library for Python. It's excellent for learning. Combined with another library such as py_ecc, which implements elliptic curves, it's possible to write Groth16 and PLONK from scratch. But it's also very useful for simply experimenting with polynomials, roots of unity, interpolation, NTTs, and many other concepts in finite fields.

To practice Rust, I'm going through the RareCode journey again. I work at the company, so I'm definitely biased, but I genuinely love this project. You learn Rust by programming in Rust: one lesson a day - about 15 minutes of your time - and in roughly three months you'll have learned a particularly challenging language. It's RareSkills-quality for about the price of a book. By the way, you can try it for free.

One of the big killers of developer upskilling is unrealistic expectations. The fundamental mismatch is that human brains cannot internalize new information as fast as many engineers would like. This leads to a cycle of aggressive learning, burnout, forgetting, and spinning wheels. Most people cannot realistically learn genuinely *new and unfamiliar* material for more than 1 hour a day, and even that is generous. Your brain needs time for new information to sit in the background while it connects to prior knowledge. Now, you can learn a lot more than 1 hour of *similar* information to what you already know, or 1 hour of *review*. For example, you could learn a new programming language in a weekend if it is close to one you already are fluent in. If you are trying to learn Rust (for example), and you already know some of it, then it is possible to be productive for 3 hours struggling with it if you have a baseline. That's because a significant amount of the three hours is spent on review. However, you cannot engage for a long time on a subject that is very far from what you already know. As an extreme example, think about trying to learn Russian if you don't even know the Cyrillic alphabet. Your brain will be cooked after 20 minutes. The problem is that learning *similar* information doesn't feel as "fast" as learning *new and unfamiliar* material. Many developers place an expectation on themselves that they should be internalizing copious amounts of *new and unfamiliar* information because the space moves so fast. But that is not realistic. By *learning* I mean really engaging with the material, not passively consuming it (of course you can passively consume for more than an hour, but that doesn't get you far).

Got someone in mind? Send them our way. We’re offering a $1,500 referral bonus

Most of the time I would encourage engineers to set aside at least 30 minutes a day to upskill technically. But there are some exceptions to this: - if you are junior (~3 years or less), you will get higher ROI by grinding at work, impressing your peers and your boss, building your reputation, and racking up clear wins. Don't spare any effort, and forget work-life balance. Any spare 30 minutes goes towards work. - if you are very senior, your technical skills might not be the bottleneck anymore, but rather your leadership, organizational, or communication skills. Spend those 30 minutes investing in those areas so you don't hit an invisible ceiling with those bottlenecks. The danger is over-relying on technical skills alone because that's what got you to where you are.

@Montyly @aave Chad move by AAVE. Congrats @Montyly

Personal update: I am joining @aave Over the past few months, I have had the chance to work closely with their engineering team. I have been impressed by how they approach security and by their attention to quality I am excited to continue helping them secure the future of DeFi

In Web3, it’s easy to chase whatever is new. A new chain, a new framework, a new narrative, a new AI workflow. But strong careers are rarely built on being early to every shiny thing. They’re built by sitting with the system long enough to understand what’s really going on. That’s where compounding starts. The deeper your technical base, the easier it becomes to move across Rust, security, protocols, ZK, and whatever comes next. Here’s what we talked In our session with Daniel Cumming about why deep technical foundations matter more than shortcuts.