Abhishek Bhave

H1B Landscape is Changing The H1B world is shifting, and I see a few trends that could shape demand: 🔹 Premium H-1B Talent – only highly specialized roles will justify sponsorship costs. 🔹 Domestic Talent – citizens, green card holders, and OPT/CPT candidates may get first priority. 🔹 Hybrid Teams – blending domestic and offshore talent will be the new norm. 🔹 Offshore/Nearshore Expansion – some projects will move to cost-effective geographies. From my view, contracting isn’t disappearing. It’s evolving. Fewer H1B-heavy roles, more domestic opportunities, and selective premium placements for specialists. 💡 How is your team preparing for these changes? #H1B #TalentStrategy #Staffing #HybridWork #ITStaffing #FutureOfWork #Cloudbridge

8

⚠️ Your "90 𝐃𝐚𝐲𝐬" are not as long as you think. Many F1 students on OPT or STEM OPT fall into a dangerous trap. They think, "𝐈 𝐡𝐚𝐯𝐞 90 𝐝𝐚𝐲𝐬 𝐨𝐟 𝐮𝐧𝐞𝐦𝐩𝐥𝐨𝐲𝐦𝐞𝐧𝐭; 𝐈 𝐡𝐚𝐯𝐞 𝐩𝐥𝐞𝐧𝐭𝐲 𝐨𝐟 𝐭𝐢𝐦𝐞 𝐭𝐨 𝐟𝐢𝐧𝐝 𝐚 𝐣𝐨𝐛." As a recruiter who has worked with over 1500+ candidates, I can tell you: The visa clock moves faster than the job market. The Reality of the 2026 Market: 𝐑𝐞𝐜𝐫𝐮𝐢𝐭𝐞𝐫 𝐇𝐞𝐬𝐢𝐭𝐚𝐭𝐢𝐨𝐧: Mentioning "OPT" often triggers extra scrutiny in the current slow market. 𝐓𝐡𝐞 𝐈𝐧𝐭𝐞𝐫𝐯𝐢𝐞𝐰 𝐋𝐚𝐠: From first contact to final round, expect 3–6 weeks. 𝐓𝐡𝐞 𝐂𝐨𝐦𝐩𝐥𝐢𝐚𝐧𝐜𝐞 𝐃𝐞𝐥𝐚𝐲: Background checks and E-Verify processing take time. 𝐓𝐡𝐞 𝐎𝐟𝐟𝐞𝐫 𝐆𝐚𝐩: Offer letters can be delayed by internal approvals. In this market, being "smart" isn't enough. You have to be strategic. The students who secure their status are the ones who stop applying randomly and start executing a structured plan. Success comes to those who: ✅ Start the hunt before the clock starts ticking. ✅ Build a targeted list of E-Verified employers. ✅ Optimize their positioning for US-specific expectations. ✅ Understand that timing + strategy = Stability. Don't wait until day 60 to get serious. If you are currently on OPT or STEM OPT and need a structured strategy to secure your future, let’s talk. 📧 𝐄𝐦𝐚𝐢𝐥: 𝐫𝐚𝐡𝐮𝐥𝐩𝐚𝐧𝐝𝐞𝐲@𝐫𝐤𝐢𝐧𝐟𝐨𝐭𝐞𝐜𝐡𝐥𝐥𝐜.𝐜𝐨𝐦  📞 𝐏𝐡𝐨𝐧𝐞: (321) 335-4904 #OPT #STEMOPT #F1Visa #InternationalStudents #USJobMarket #CareerStrategy #Hiring2026 #JobSearchTips

25

1 Comment

I'm curious why big companies still use H1b visas while there is L visa. Advantages for a company: - No quotas, i.e. it's possible to move anyone any time. - L visa doesn't allow switching jobs (as opposed to H1b that does allow transfer). A company has to be international (i.e. to have offices in 3+ countries) and an employee has to spent 1 year abroad before they can be moved. Obviously H1b is more flexible for an employee. But, on H1b your close relatives neither get work authorization nor SSN. A dependable of the L visa holder gets working permit right away. Saying that, I guess, the new giant fee for H1b visa will just move big players towards using L. For small companies... I bet more companies might want to open offices overseas. UP: after I posted that I learned that L visa cannot be extended for more than 5-7 years while H1b can be extended for longer (if you get your GC I-130 form approved). So, H1b works for those who need to wait for 10+ years for GC as a permanent move solution and L doesn't. IIUC only Indian and Chinese citizens have such big wait times.

68

17 Comments

I'm an SDE working at AWS with 10 years of experience in software engineering. I've spent the last 1 year of my life experimenting, building and working on AI infra at AWS. From my experience in just the last 8-10 months, I can tell you for a fact that: Building a prototype with RAG is easy. Quickly testing out another agentic solution is easy. Training a small ML model on a clean sample dataset is easy. Deploying an ML service that works in the happy path is easy. What is hard is doing all of this with: –  messy real-world data – strict latency and availability targets – compliance, security, and privacy constraints – fixed budgets and unforgiving timelines and still creating meaningful business impact. Here are a few things I have learned about building with AI: 1. Start from constraints Before you make any infra choices, ask: how fast does this need to respond, how much can it cost, how much traffic will it see, what rules do we have to follow. The answers shape the design. 2. Simple designs are great Simple, predictable systems that keep working under load, during deployments, and on bad days beat any clever stack that breaks in production. 3. Prepare for life after launch How will you debug weird outputs, roll back a bad change, add a new data source, or explain behaviour to a stakeholder. If you have answers to that, you are building a system, not a one-time demo. Learning all of this takes time. 'Upskilling' for AI is not a 1 month crash course. Being an 'AI engineer' is realistically quite a long journey, compared to what most gurus market and promise online!

109

22 Comments

Watching the recent Amazon layoffs unfold reminded me that even the strongest performers aren't immune to company restructuring. If you're one of the folks affected, here's what helped me/others I know in the past: 1. This is not a reflection of your skills or worth. Companies make decisions based on many factors beyond individual contributions. 2. Take some time to get your head in the right space. It could be a day, a week or more if you can swing it. Go for walks, spend time with family, binge on shows - whatever helps. The best decisions are made with a clear mind. 3. Start hitting up your network. The market is tough but opportunities are out there. Let people know you're looking. Recruiters are looking to hire fast, and a referral is faster than sifting through job resumes. 4. Update your LinkedIn profile and resume. Make sure your headline reflects what you want to do next, not just your last title. Highlight your achievements and the impact you've made in your roles. 5. Start a side project or contribute to open source. It keeps your skills sharp and shows potential employers your passion and initiative. That idea you've been sitting on? Now's the time to build it. 6. Find your community. Whether it's tech meetups, online forums, or Discord servers - being around other builders and job seekers helps. You're not alone in this. 7. Don't give up. Statistically, people who have been laid off often find better opportunities than their previous roles. Stay positive and persistent. That's it for now. If you have any questions, or just want to chat, feel free to reach out. Good luck out there! #SoftwareEngineering #CareerAdvice #JobSearch #Layoffs #TechCommunity

5

2 Comments

What I would do to beat behavioral interviews as a Software Engineer at Amazon Interviews usually break down into two parts: behavioral and functional. Functional covers the job family. For software engineers, that’s coding, algorithms, and system design. For PMs, TPMs, designers, etc., it’s domain-specific skills. Behavioral is where a lot of candidates underestimate the prep and think they can just improvise during the interview. That's a bad idea because at big companies, behavioral interviews carry serious weight. If you don't come prepared to demonstrate how you work in a fashion that aligns with the company's core values, it will be nearly impossible to land an offer even if you passed the functional interview with flying colors. Here’s how I prepare for behavioral interviews: 1. Know the company’s values. At Amazon, they’re the Leadership Principles. Google, Meta, and others have similar ones. Every question ties back to them. 2. Research common behavioral questions. Each company has patterns in what they ask, and they are typically designed to extract signals from candidates to match their core values. 3. Prepare your narratives using STAR (Situation, Task, Actions, Results). One narrative can often map to multiple principles. You don’t need 16 unique narratives. 4. Go deep. Interviewers will probe for follow-ups. Think ahead: “Why did I choose solution A over solution B? What trade-offs did I consider? What would I do differently?” 5. Practice relentlessly. Simulate the real thing and get feedback. If you can afford to hire an industry professional (typically costs between $100-$250 per hour), great. If not, I built PrepRoom.io to give realistic, voice-based mock interviews and structured feedback (free to use now, feedback appreciated!). And here’s the thing - don’t wait until you’re actively job hunting. Hone your interview skills regularly, so you’re ready when opportunity shows up. Comment below: How often do you practice your interview skills, even when you’re not looking? --- Hello, I’m Charles, and I am committed to share my two cents on career-related topics for 100 consecutive days. In these uncertain times, I hope to support those facing layoffs or career challenges. Follow me, and let's navigate this together! (56/100) #InterviewTips #JobSearch #CareerGrowth #BehavioralInterviews #MockInterviews

23

8 Comments

I’ve interviewed 30+ Sr. SDEs going for L5/L6 level roles at Meta, Google & Amazon. A lot of them failed the database question in the SD round. The mistake? They name the technology before they understand the workload. If you want the 1Cr+ total compensation, you have to stop thinking like a coder and start thinking like an architect. Here is the exact 4-step framework I use to map any workload to the perfect tech stack: Step 1: Label the workload Before naming any database, answer 3 questions in your head. 1. Is this OLTP vs Analytics vs Search/Logs 2. Is traffic read heavy, write heavy, or balanced 3. Do we need strong consistency or is eventual consistency fine Once you know this, the search space is already much smaller. Step 2: Choose the data model family Pick a model instead of a product: - Relational tables - Document - Key value - Columnar analytics - Time series - Search index Quick rules of thumb: - Complex relationships, transactions, money flows -> relational - Flexible JSON-like records, changing schema -> document - Simple id lookups, very hot keys -> key value - Heavy reporting on large history -> columnar analytics - Metrics or events over time -> time series - Free text search, fuzzy matching, ranking -> search index Step 3: Map to real technologies Now you can safely talk about specific tools. 1. Classic relational OLTP Use for: orders, payments, bookings, user accounts Tech: Postgres, MySQL 2. Distributed SQL for high scale or multi region Use for: same as above, but with very high scale or multi region writes Tech: CockroachDB, YugabyteDB, TiDB 3. Key value and caching layer Use for: sessions, feature flags, counters, rate limits Tech: Redis, Memcached 4. Document store Use for: user profiles, settings, content documents Tech: MongoDB, Couchbase 5. Wide column store Use for: large scale write heavy workloads, telemetry, big sparse tables Tech: Cassandra, HBase 6. Columnar analytics and OLAP Use for: BI dashboards, offline analytics, aggregates on billions of rows Tech: ClickHouse, other columnar warehouses 7. Time series and metrics Use for: monitoring metrics, IoT sensor data, time based events Tech: TimescaleDB, InfluxDB, Prometheus 8. Logs and search Use for: application logs, product search, log search Tech: Elasticsearch, OpenSearch Step 4: Internal decision flow, you can sketch In the interview, you can even draw a tiny flow like this: 1. Is it core transactional data with strong consistency needs    -> Yes: start with Postgres or MySQL    -> Need multi region or very high scale: move to CockroachDB / Yugabyte / TiDB 2. Is it analytics or reporting over large history    -> Yes: replicate into ClickHouse style columnar store 3. Is it metrics or time series    -> Yes: Timescale / Influx / Prometheus 4. Is it full text search or logs    -> Yes: Elasticsearch 5. Do we need very low latency for a small working set    -> Yes: add Redis cache in front of the chosen primary store

410

27 Comments

Sit in for a Dev role at Google, they grill you on DSA for three or more rounds. Sit in for a Dev role at Amazon, Uber, Microsoft same story. Then you join the job and nobody is asking you to invert a binary tree to save the quarter. The interviews and the actual work often feel like two different worlds. You can call it a broken process, but right now it is still the process. So the game is simple: learn how to perform on DSA without losing your mind or cheating your way through it. If I had to start over from zero, this is how I would train for DSA. Before we dive in - here’s a free DSA mock interview tool we built at Layrs. If you want structured practice with real feedback, go to layrs.me/interviews. 1. Win the first 30 minutes Most coding rounds are decided early. Practice solving the first 2 to 3 problems of a set in under 30 minutes. You want: - One easy problem in under 5 to 7 minutes - One more easy or easy medium in the next 10 minutes - One solid medium before the clock hits 30 You are training speed and remaining calm.  2. Train debugging as a real skill Guard your brain from “submit and pray.” For every problem you solve, spend a few minutes asking: - Which edge cases can break this - How I would find a failing test if the judge says wrong answer - What pattern of mistake I made People who climb rankings are not guessing more. They are fixing faster.  3. Build a focused toolbox. Pick a small set of topics and get very good at them: - Arrays, hashes, two pointers, binary search - Classic DP and a bit of bitmask DP - Trees and graphs with BFS and DFS - Basic greedy and prefix sums You can add fancy stuff later.  First make sure you can crush the common 80 percent.  4. Use templates and a repeatable system Have ready skeletons for: - DFS and BFS - Common DP layouts - Binary search on answer - Graph adjacency lists The less time you spend fighting syntax, the more time you have to think.  5. Measure progress. Instead of chasing green days, track: - Problems solved under 20 or 30 minutes - Problems you could fully explain to a friend - Topics where you still panic Interviews are not going to change tomorrow. Your approach can change tonight.

26

Once you’ve worked long enough as an SWE (spent 8 years like me), facing every disaster you can imagine and been humbled by every sort of problem at companies like Meta, Google, LinkedIn, and Amazon. As a Software Engineer, I want to tell you that your fundamentals set the tone for your career. It’s never about how many frameworks you know. → It has and always will be about how quickly you adapt to new ones. It’s never about writing the most lines of code. → It has and always will be about writing code that’s easy to read, debug, and maintain. It’s never about building the fanciest system from day one. → It has and always will be about solving the actual problem and scaling only when needed. It’s never about acing every interview or online test. → It has and always will be about solving problems nobody saw coming—while under pressure. It’s never about holding a prestigious job title. → It has and always will be about the impact you create for your team and customers. It’s never about being the smartest person in the room. → It has and always will be about helping others get better alongside you. It’s never about knowing every tool out there. → It has and always will be about picking the right tool for the job, even if it’s not trendy. It’s never about chasing 100% code coverage. → It has and always will be about understanding what’s truly critical to test. It’s never about running after every shiny new technology. → It has and always will be about mastering the basics that never go out of style. It’s never about perfection on the first try. → It has and always will be about getting feedback, iterating, and learning as you go. No matter where you work or how senior you get, the engineers who last are the ones who stay grounded in the basics, solve real problems, and help others win. That’s what sets you apart.

195

24 Comments

This month, I turn 9 at Amazon. In tech, that’s often considered a mistake. The common advice is clear: Switch companies every few years. Reset compensation. Collect titles. Keep moving. And there’s truth in that. But something interesting happens when you stay long enough. You stop just building systems and start seeing the consequences of them. You watch an architecture decision you made three years ago struggle under 10× scale. You see which shortcuts actually cost the organization later. You inherit the operational pain of things you once argued were “good enough.” That kind of feedback loop is rare. Most career moves reset the clock before those lessons arrive. There are good reasons to leave a company. But there’s also a different kind of growth that only shows up with time: Trust compounds. Scope expands quietly. And the systems you helped build eventually teach you what you didn’t know when you built them. Nine years isn’t loyalty. It’s staying long enough to see which of your decisions survive contact with time.

262

12 Comments

Etched just raised ~$500M in a major funding round (valued at ~$5B) to accelerate its custom AI silicon — a solid alternative to Nvidia-centric stacks. https://lnkd.in/dttxqJPJ

6

What does it take to serve agentic workloads on trillion-parameter models at 400 tokens per second per user — without trading throughput for latency? The NVIDIA Vera Rubin platform pairs Vera Rubin NVL72 with NVIDIA Groq 3 LPX to deliver low latency on trillion-parameter MoE models with 400K-token context with a 35x higher throughput per megawatt. Learn how the deterministic LPU chip-to-chip (C2C) fabric and extreme co-design address agentic AI's scale-up challenges. ➡️ https://bit.ly/4uenrHH

2

In 2022, Coinbase rescinded accepted offers. In 2023, several companies delayed joining dates or pulled back campus hiring. Today, we all woke up to the news of Oracle reportedly revoking offers across IITs and NITs. This is not the first time students have been left in this situation, but most campus placement systems are still built as if an accepted offer is final from both sides. That is where the problem begins. In many colleges, once a student accepts an offer, they are locked out of the placement process. The logic is understandable. It gives other students a fair chance, prevents offer hoarding, and helps placement cells manage limited opportunities. But when a company revokes an offer, that same rule leaves the student with almost no safety net. They followed the process. They accepted the offer. They could not sit for other companies. Their family may have already started planning around that job. Then suddenly, one month before joining, they are back in the market when most campus hiring has already moved on. This is not just an individual company issue. It is a structural gap. Placement cells need to rethink how they protect students in such cases. Maybe students with revoked offers should be allowed back into placements immediately. Maybe colleges need a backup hiring window. Maybe companies should face stricter accountability before making campus offers they may not honor. I understand the fairness argument. But fairness should not only mean “one student, one offer.” Fairness should also mean that when a company backs out, the student is not left carrying the entire risk alone. For a company, this may be a hiring decision. For a student, it can be the first major shock of their career which can set them back anywhere from 6-12 months, which is a big disadvantage.

95

5 Comments

H-1B Visa Fee Hike: A Game-Changer for Global Talent? Donald Trump has signed a proclamation introducing a mandatory $100,000 fee for H-1B worker visa applications, a drastic increase from the current $1,000-$5,000 range. This move is framed as a way to ensure companies hire only the most highly skilled foreign workers who can't be replaced by American talent. The massive fee could make it financially impossible for many companies, especially startups and mid-sized firms, to hire foreign professionals. * What are the long-term implications for the U.S. tech industry? * Will this drive innovation and talent to other countries? * How will this impact global supply chains and skill diversity in the workforce? The conversation around global talent is more critical than ever. Let's discuss. #H1Bvisa #Immigration #TechIndustry #USbusiness #GlobalTalent #Innovation #Workforce

4

This Amazon layoff has been really hard to cope with. 17 years, and you just get tossed aside like you have no worth anymore and didn't contribute to anything. I've always been really bad about imposter syndrome. Everything I know as a software engineer I learned from just doing on the job. Anything I needed to know, I learned on the fly. But this doesn't translate well in my head. "Oh, you're not a real software engineer. You didn't go to school for it. You didn't learn all the CS basics people poke at in tech screens. How will you end up even getting interviews, much less getting to the point you can really prove what you know and can do?" it doesn't help that I kind of got pigeonholed into the space I was working in Amazon, very specific use case, for more than a decade. I feel like I've forgotten everything else, and don't know where to find a position relatively within the same space. I'm still motoring along, have a few calls lined up, but it still doesn't mean my confidence and anxiety aren't still playing a big factor in things. I don't know what to do about it. I just feel broken. Does anybody else who feels the same way have any tips, hints, or even just a word of encouragement? I could definitely use a little.

1,690

109 Comments

A junior engineer at Amazon nearly burned herself out and put her career at risk, doing the work of two people alone, thinking it would get her promoted… without ever saying a word. Back when I was at Amazon, I noticed a strange pattern with a junior; she was quietly taking on more and more work, often beyond work hours, without ever bringing it up. You see, some people just won’t ask for what they deserve; they just keep giving until the other person notices or they break. And that’s exactly what happened here… The Junior engineer was quiet, consistent, and brilliant. For months:  - she would fix issues that went unnoticed  - she would unblock teammates outside of work hours - she would always show up early and was the last one to go… Even after doing the work of 2 people, she was always calm and present even though she was struggling and burning out. I think being young, she herself didn’t realize that all of this extra effort was costing her everything, and her manager didn’t even know… She just kept hoping that one day someone would finally see how much she was giving. But nobody did, because she never spoke up for herself. When I asked her about her promotion, she finally opened up and admitted she didn’t know what else to do. We talked, and I told her to set up a meeting with her manager and lay out everything she was handling, all the extra hours, and all the work nobody saw. When she finally did, her manager was shocked. He said he had no idea she was taking on so much. He genuinely thought she was comfortable and happy with her workload. That single meeting made all the difference in her career. After that conversation, they made a plan, mapped her progress, and set clear expectations for the next few months. And just like that, in the next review cycle, she got the promotion she deserved because she finally found the courage to ask for it. This is what I keep seeing in our industry: Too many of us live that same story. – we over-deliver. – we under-communicate. – and we wait for a signal. But sometimes,   the signal you’re waiting for   is the one you were meant to give. Sometimes it’s not just about work.   It’s about the things we leave unsaid,   hoping they’ll be understood. If this made you think of someone,   maybe it’s time to tell them to speak up. Silence only feels safe…   until it costs too much.

213

16 Comments

🚨Despite the new $100,000 fee per H-1B visa, Jensen Huang confirmed that NVIDIA will continue sponsoring visas & cover all costs. His reason? “Immigration is central to America’s ability to lead in technology and ideas.” It’s a bold move and a powerful message. Leadership isn’t about cutting costs. It’s about standing by what drives innovation. If you’re a professional aiming for roles in AI or advanced tech, this signals two things: ➝ Top companies will always pay for the best talent, no matter the policy climate. ➝ Your value grows when your skills travel well, when you can adapt, collaborate and build across borders. For founders and leaders, it’s a clear reminder: Don’t let policies decide how you hire or grow. Let your vision lead the way. ➡️Take for students: Build expertise that transcends location. The world’s best companies look for capability, not just geography. 1️⃣Master one domain deeply, go beyond your syllabus. 2️⃣Build real projects and show your work online. 3️⃣Collaborate globally through hackathons or open-source. 4️⃣Focus on solving real problems, not chasing buzzwords. 5️⃣Communicate and teach what you learn, it builds clarity and credibility. Remember: Your strong skills are your passport. They can open doors anywhere in the world.

33

3 Comments

Imagine you're sitting in an interview for an L5/L6 Software Engineer role at Google, and you're asked this question: "Design Uber for a metro city." Here's how I would break it down if I were the one being interviewed. There are lots of aspects we can cover here, but I'll only cover the high-level design. Btw, if you’re preparing for system design/coding interviews, check out our mock interview tool. You can use it for free here: https://lnkd.in/gpCn7t2T  1. Requirements The classic pain points in ride-hailing apps: - Matching is slow because the system can't efficiently find the nearest available driver from a pool of hundreds of thousands in real time - Consistency is fragile because concurrent ride requests can accidentally assign the same driver to multiple riders - Location tracking is expensive because millions of driver pings per minute overwhelm traditional databases Functional requirements - Rider inputs a start location and destination, gets a fare estimate and ETA - Rider requests a ride based on that estimate and gets matched with a nearby available driver in real time - Driver receives a ride request notification and can accept or deny it - Driver navigates to the pickup location, then to the destination, updating status along the way - Drivers periodically send their current location so the system knows where they are Out of scope (worth mentioning to show product sense) - Multiple car types (UberXL, Black, etc.) , we'll assume one tier - Ratings and reviews for drivers and riders - Scheduled rides in advance - Surge pricing algorithms - Payment processing Non-functional requirements - Low-latency matching: under 1 minute from ride request to driver match (or a failure message) - Consistency of matching: a ride is matched to exactly one driver, and a driver is never double-booked across rides - High availability outside of matching: the rest of the system (fare estimates, location updates, navigation) should be up at all times - High throughput in surges: the system must handle hundreds of thousands of requests within a region during festivals or special events Rest of the breakdown: https://lnkd.in/gieR-EqR

276

12 Comments

🚨 60 day grace period for H1B workers who lost their job is no more Most software engineers moving to the US for a job do so with a H1B visa – you can get a L1 (but then your stay is tied to the company) or an O1 (but that's fairly restrictive). In the past, in case you lose your job, you could stay 60 days in the country to find another job. After that, you have to leave the country but can still come back if you get a new job and get your H1B transferred (e.g. you don't need to go through a lottery again) The issue now seems to be that the 60 days grace period is gone[1], so essentially you have to immediately leave the country once you're laid off. Being an immigrant always come with some uncertainty on your situation but this wasn't too much a problem in the past because: - Tech companies were not doing laid off - The job market was in favor of engineers – it was easy to just find another job - While 60 days is kind of short, it was still enough in many cases (or at least to sort part of your personal situation) This is a pretty big blow for the industry – while some (especially younger folks?) will still move to the US, less and less will be willing to do (especially if you have no short path to a green card) The situation is also even worst for people in the current situation that may be overstaying because their 60 days grace period was suddenly gone and may not be able to transfer their H1B visa after 🙁

73

20 Comments