Activity
2K followers
-
Huy Khieu shared thisDear Santa, I've been a good product builder this year. I promise. 🎅 All I want for Christmas is simple: our first paid customer for Rework Times. Here's what we've built: → A remote job board laser-focused on Vietnam & Asian talent → 3,000+ monthly visitors who actually care → 11.6k Threads followers (we're kind of a big deal there) → 879 email subscribers who open our emails → 221 LinkedIn followers (still small) Here's what we need: → 1 company brave enough to be customer #1 → Your help spreading the word to companies hiring in Asia Santa, I know you're busy, but could you help the right company discover we've built something special? We've got the audience, the traffic, and the passion. Now we just need that one customer to believe in us first. Think you can slide that down the chimney? 🎄 P.S. - If you're reading this and hiring remote workers in Asia, consider this your Christmas sign. DM me, and I'll throw in extra Christmas karma. #RemoteWork #StartupLife #ChristmasWishes #Vietnam #AsianTalent #ReworkTimes
-
-
Huy Khieu reposted thisHuy Khieu reposted this$13M says AI agents are real. Mastra just became the third-fastest-growing JS framework (and ai agent builder) ever - and Sam Bhagwat is quietly rewriting how dev teams ship AI. I was talking to Sam the other day, who is an absolute force in the agent space btw Mastra, his new TypeScript framework, just raised $13M from 120+ top Silicon Valley investors. It is now the third-fastest-growing JS framework ever. Replit Agent 3 builds Mastra agents at scale. SoftBank announced a Mastra-built product on stage. Marsh McLennan rolled out an agentic search tool to 75,000 employees. And yes, Sam brought back much of the original Gatsby (A Netlify Company) team. My 5 takeaways from our interview 👇 1) Agents are the platform shift - On a 10-person team, the 2–3 developers who master agents become **2–3x more productive. - YC is ground zero for vertical agents. It feels like the 1996 internet moment all over again. 2) Orchestration beats models - Mastra focuses on workflows, tools, memory, evals, tracing, and guardrails** so agents don’t just work, they actually ship. 3) B2D that lands enterprise - Win developer love with open source, prove value inside orgs, then formalize with cloud or support. - Treat cloud PMF as its own motion. Don’t outsource it too early. Figure it out first. 4) Where agents win right now - High-volume, repeatable, paperwork-heavy operations like **insurance claims, finance, and healthcare. - Clear input, transform, output. That’s the agent sweet spot. 5) Careers in the agent era - Traditional recruiting channels are being flooded by AI automation. Hire through trusted networks. - It’s a learn season, not an earn season. Lean into AI now and compound later. PS: Sam and team printed 70k+ copies of their book Principles of Building AI Agents_ and distributed 100k digital copies. Want one? Comment BOOK and I think :) he will send you a copy if he still has one lying around somewhere. If you write TypeScript or run a team that needs leverage, this episode is worth your commute. Subscribe for more blunt, no-fluff conversations with people who are actually shipping AI at the homebase dot ai or check out our youtube channel called the AI Chopping Block!
-
Huy Khieu reposted thisHuy Khieu reposted thisMany people have written us to praise our work at The Homebase AI and our podcast The AI Chopping Block. “How are you guys doing it?” “What is your secret?” “Any advice?” The truth is even with all the support, we constantly question ourselves if people like it enough and the reality is we haven’t figured it out too. So we’re changing our approach by going more personal. More "Behind-The-Scenes. Sharing more lessons learned, whether they are good or bad. Each week we set a new simple rule for ourselves: focus on one task, measure it, and by Friday share one lesson or result. What worked or what needs to be dropped? (both equally important) If something failed, is there other way to test it? Did we really try them all? What about AI tools? Did things got better or worse? These are the questions anyone building something is asking. One thing I love about the AI community is how people desire to share. So every Friday, I’ll post our lessons: wins, mistakes, takeaways, and everything in between. I will not only share just the professional side, but my personal one too. Because being a professional also means honoring the human side. We all have thoughts and feelings that show up in our work. What to expect in these Friday posts: - AI tools we tested or AI insight learned during the week. - Challenges and learnings from working in a remote, multicultural team. - Personal reflections and experiences. My dreams. My projects. It is time for all of us to build in public. Real > Perfect. The people creating amazing stuff are as humans as all of us. Share me something you want and let's be more open together! 🙌 ♻️ Repost this to inspire others and follow Isabel Quiñones for more. :)
-
Huy Khieu shared thisI built a LinkedIn content library that turns viral Reddit posts into LinkedIn drafts I got early access to Softr Workflows from Softr and decided to build something I actually needed: an AI content assistant that does the heavy lifting for me. Here's how I built it 👇 ✅ Step 1: Content Discovery - Submit any Reddit URL through a simple form - No manual copying or formatting needed ✅ Step 2: Crawling (Apify) - Automatically extracts the full post content ✅ Step 3: Strategic Analysis (OpenAI GPT-4) - Identifies the core story and key insights - Analyzes target audiences and pain points - Determines what translates well to LinkedIn - Finds natural engagement opportunities ✅ Step 4: Content Creation (Claude Sonnet 4) - Generates 2 unique LinkedIn drafts - Each with different angles and audiences - Maintains an authentic, professional tone - Includes hooks, hashtags, and conversation starters ✅ Step 5: Organized Storage (Softr Database) - Stores everything in one dashboard - Original content, analysis, and drafts - Easy to review, edit, and publish What surprised me most: All of this runs natively inside Softr Workflows. No Zapier, no Make, no n8n, no juggling multiple platforms. Just one workflow connecting everything automatically. The result? From Reddit URL to polished LinkedIn drafts in under 2 minutes, with everything organized in one place. If you want to try it out, Softr’s giving away 500 free Workflow Actions with their free plan. Comment “SOFTR” and I’ll send you the link 🚀 #SoftrWorkflows
-
Huy Khieu shared thisShould I move from Make to n8n? As an automation freelancer who builds workflows for my business and clients, I've been a Make user for a while. This week, I tried n8n with Hostinger for the first time. And wow, it's way cheaper than Make's monthly plan. n8n is powerful if you know JavaScript, and Make's interface is definitely simpler to work with. But here's my issue: I'm constantly hitting operation limits and paying extra beyond the $34.12/month Core plan. Hostinger at $12.99/month seems like a way better deal, right? And something people don't mention much, Make has been getting slower lately, especially with complex modules like OpenAI.
-
Huy Khieu shared thisThis man built an n8n workflow that automatically finds clients and made him the first $227 This workflow generates client revenue by automatically finding and reaching out to e-commerce stores with slow websites. ⚙️ How It Works The workflow runs in four automated steps: 1. Input: Starts with a list of e-commerce websites 2. Speed Test: Uses RapidAPI's Website Speed Test API to check each site's performance metrics 3. Filter: An IF node filters sites with load times over 3 seconds 4. Outreach: Automatically generates a performance report and sends personalized emails to business owners The system identifies the problem, creates the report, and initiates the conversation—completely hands-off. 💼 The Business Model The workflow handles lead generation and initial outreach. When clients respond and deals close, freelancers execute the actual speed optimization work for a 20% fee. The system essentially acts as an automated project manager, connecting clients with solutions while keeping 80% of the revenue. 💡 Why This Matters This isn't about $227. It's about proving that automation can create real business opportunities when targeting genuine pain points. The workflow runs 24/7, finding qualified leads while focusing on higher-value activities. Source: https://lnkd.in/g4Wg4nhW
-
Huy Khieu shared thisI spent 2 weeks testing 10 AI tools for LinkedIn writing. Here's what I actually found (with honest takes + pricing) 👇 I'm obsessed with finding tools that make LinkedIn writing faster without losing authenticity. So I tested every AI tool I could find. Here's the breakdown: EasyGen ($59.99/month) - Founded by Ruben Hassid (688k followers - major AI thought leader) - 100% focused on LinkedIn - Standout feature: Image-to-LinkedIn-post (unique!) - Premium pricing, but laser-focused on LinkedIn creators Typefully ($12/month - yearly only) - Very famous on X - Used by top X creators (that credibility matters) - Super intuitive interface - Best value for multi-platform creators Supergrow ($19/month) - Pure LinkedIn focus - Tons of testimonials (social proof checks out) - Their landing page alone shows they get marketing - Solid mid-tier option RedactAI ($19.90/month) - LinkedIn-only tool - Killer feature: Curates viral + competitor posts for inspiration - Free tools available (great for testing) - Smart approach to content research MagicPost ($39/month) - LinkedIn specialist with strong testimonials - 9k followers = decent traction - Higher price point, positioned as premium Bright (Free!) - LinkedIn content management - No AI yet, but excellent for organizing ideas - Free = perfect for beginners or budget-conscious creators Also Worth Knowing: Dottypost ($15/month) – Affordable LinkedIn carousel specialist Publer ($5/month/seat) – Multi-platform, great for teams Socialsonic ($20/month) – Backed by Writesonic, LinkedIn-focused Typegrow ($29/month) – LinkedIn scheduling and growth features The truth: No tool writes viral posts FOR you. But the right tool can: - Cut your writing time in half - Give you content ideas when you're stuck - Help you stay consistent What's your experience with these tools? Anyone using something I missed?
-
Huy Khieu shared this🟡 Remote Jobs Weekly Updates No.7 Check out the latest remote opportunities we’ve gathered for you this week. ↓ 1. SOFTWARE DEVELOPMENT ✢ Senior/Staff DeFi Engineer at Ethena Labs ✢ Backend Developer (C++) at Magic Media ✢ Strong Junior Unity Developer at Growe ✢ Freelance AI Red Team Engineer at Mindrift ✢ Senior Backend Developer (Python/Django) at Merchway 2. MARKETING & SALES ✢ Project Manager at Spiralyze ✢ Office Administrator at Employment Hero ✢ Payroll Customer Success Manager at Remote ✢ Marketing Internship at Glints ✢ Conversion Rate Optimization Expert at Hustler Marketing 3. DESIGN & CREATIVE ✢ Senior Growth Designer at Circle ✢ Senior Presentation Designer (Remote, Contract) at INFUSE ✢ UX/UI Designer at Piktochart ✢ 3D Design & Implementation Specialist at CLO Virtual Fashion 4. AI TRAINING & LABELING ✢ Senior Security Engineer - Enterprise AI Security at GitLab ✢ Crypto Quantitative Analyst at Token Metrics ✢ Freelance Chemistry Expert - AI Trainer at Mindrift ✢ Freelance Mathematics Expert - AI Trainer at Mindrift Follow Rework Times for weekly remote job updates!
-
Huy Khieu liked thisHuy Khieu liked thisTata Consultancy Services reported its strongest hiring quarter in four years while cutting 12,000 US-based positions. Nearly half of its revenue still comes from US clients, yet delivery work appears to be consolidating in lower-cost regions rather than expanding locally. The signal isn't weaker demand. It's a shift in where the work gets done.
-
Huy Khieu liked thisHuy Khieu liked thisForward Deployed Engineer postings went from 643 to 5,330 in twelve months. A 729% jump (Indeed, Apr 2025 to Apr 2026). Everyone is still arguing about whether AI kills engineering jobs. Meanwhile the market already answered. It didn't kill the job. It split it. The roles getting crushed are the ones that were always a proxy for "can write code." Those are now a commodity, because the model writes the code. The roles going vertical are the ones AI can't do alone: → Forward Deployed Engineer: turns a demo into a production system inside a customer's messy stack. Comp clustering at $300K to $550K TC, principal roles clearing $1M. → Applied AI Engineer: stands up RAG plus evals plus agents that actually hold up in prod. → AI Evals Engineer: the person who can prove the agent works before it embarrasses you. Notice what these have in common. None of them are "coder." They're all judgment plus customer contact plus ownership. I'm hiring for exactly this profile right now. The bar isn't "knows Python." Roughly 41% of AI engineers now spend over 30% of their time customer-facing (Stack Overflow 2026). The ones who can't sit in front of a customer and decompose a vague problem are getting screened out, no matter how clean their code is.
-
Huy Khieu liked thisHuy Khieu liked thisHow to become a $1M forward-deployed engineer (FDE) in 30 days (and what FDE clearly means): - What an FDE is and why the role EXPLODED - The 3 stages of the job (business reality, judgment, building) - The exact 30 day roadmap to become one - What the top roles actually pay Full ep: https://lnkd.in/eE8BzCTV Enjoy. Even if you never want the job, this teaches you how to actually implement AI agents in your own business, which workflows to automate, where AI belongs, and where it doesn't.
-
Huy Khieu reacted on thisHuy Khieu reacted on thisHere's to the visionary non-technical founders trying to hire a "Founding AI Engineer" to build their entire agentic platform on a junior dev budget. There's just two small problems with that. Problem one: the technical reality. Building a AI agent that has any semblance of moat isn't just making a quick API call to OpenAI anymore. It requires deep expertise in state management, vector databases, retrieval-augmented generation (RAG), tool-calling, and mitigating loops. That's why the engineers who actually know how to build reliable agentic systems (not just wrappers) are fielding $500k+ base salaries from top AI labs and well-funded startups. $85k isn't a lowball for that kind of talent, it's a whole different sport. Problem two: the leverage. The technical folks capable of building an agent from scratch don't need your idea. If they can build an agent that works and solves a high-value problem without breaking in production, they are sitting on a goldmine and can build and launch it themselves over a weekend. They don't need to do 100% of the work for 0.5% equity while you act as the ideas guy. So when you pair ChatGPT-wrapper budgets with AGI-level expectations, you get… maybe a demo that breaks in production and an agent that'll be obsolete in 6 months. Your options are simple: → Pay the market premium (300k+) for an actual AI engineer who knows how to do the above. → Give up 50% equity to make them an actual co-founder with real skin in the game. → Roll up your sleeves and use no-code agent builders (like Langflow, Flowise, or Voiceflow) to hack together the MVP yourself to prove traction, then do one of the above. What's NOT an option: Offering 0.5%, an $85k salary, and expecting a single engineer to build the autonomous workforce of the future for you. Ideas are cheap. Good AI engineers are more expensive than you think.
-
Huy Khieu liked thisHuy Khieu liked thisA profitable company just put $100M into a brand-new AI startup - and we're helping them hire across the board right now. Here's the setup: the parent company is a profitable, team-owned platform serving 8+ markets. No outside investors. They funded this new AI venture internally with $100M and capital ready behind it. US HQ, global team from day one, run as an independent startup - not a side project. The mission: 5 billion people use email, notes, and task apps every day that aren't AI-native. This team is building the proactive assistant that fixes that - real task completion, persistent context, workflows that actually finish. A few of the remote roles open (this is not the full list - they're hiring across the board): → Principal ML Engineer The deep technical authority on training, inference, evaluation, and GPU infrastructure. You set the standard for how ML systems get built. Distributed training, inference optimization, zero-to-one ownership. Hands-on, not a manager seat. → Head of Engineering Turn technical direction into shipped product. Structure the workstreams, keep velocity high, and build the engineering team across the US, Europe, and Asia. Strong architecture background required. → Full-Stack Engineer, AI Systems Build the product layer on top of the models — agent workflows that plan, use tools, fail, and recover. Next.js, Python, LLMs, RAG. Ship AI features that go beyond chat into real goal-driven workflows. → Desktop Engineer Own the desktop app end to end. Electron, TypeScript, React, real-time collaboration, native OS behavior. Make it fast, stable, and genuinely magical to use. Most of these are $150K+, several well above and fully remote. Top-of-market comp, 401k match, full health/vision/dental, sponsorship available. Some seats they want in the US. Many can be based anywhere. And this is one of several clients I'm hiring for right now - enterprises and startups both, heavy on the AI side, most of them six figures. If you're an engineer who wants to build AI that actually ships - DM me. If you know someone - tag them below. P.S. Separately, I've also got a frontend-focused remote Software Engineer role open at a well-funded travel tech company. $150K–$250K, remote. Bar is high - think FAANG-level interview - and you'll need real backend chops too, specifically building and setting up RAG systems. If that's you, my DMs are open.
-
Huy Khieu liked thisHuy Khieu liked thisMeta's July layoffs appear to have used algorithmic selection criteria that targeted employees on approved leave status. The selection mechanism suggests workforce reductions may now operate independently of performance evaluation cycles or manager discretion. Twenty-six affected workers challenged the timing in federal court, arguing the algorithmic process created disparate impact, but a judge ruled job loss doesn't constitute irreparable harm sufficient to block the cuts. The case indicates large platform companies are moving reduction decisions from human managers to automated systems that optimize for labor cost variables rather than individual circumstances.
-
Huy Khieu liked thisHuy Khieu liked this120k+ views on a post that will split every engineering org into two camps in 2026. Sam Lambert, CEO of PlanetScale, shared that they passed on a talented engineer who was "ethically opposed to using AI". His question: how is that still a viable stance. My 2 cents is that it's not. → PlanetScale probably ships at a pace that assumes AI-assisted workflows. A new hire refusing AI isn't a moral stance, it's a productivity gap waiting to happen. → "Ethically opposed" is doing a lot of work in that sentence. What's wrong with using AI (a tool) to be better at your job? Ethics should be about how you use a tool, not whether you do. Refusing to touch AI on principle as a SWE in 2026 isn't integrity, it's opting out of the job.
-
Huy Khieu liked thisHuy Khieu liked thisI need to hire 100 people in the next 12 months. Tenex is a rocket ship. Unlike anything i've seen. Our biggest bottleneck is keeping up hiring with the tidal wave of demand we're seeing. Which takes me to the first role I need to fill: Director of People Ops. We need a hands-on, foundational leader who will build and scale the People function as the company grows from ~15 to 100 employees and beyond. Must haves: - 5–10 years in HR/People Ops, ideally across multiple scaling stages (50–250 employees) - Prior experience as a Head of, Director, or Senior Manager of People in a high-growth SaaS or startup. - Pumped to own all aspects of People Ops, Employee Experience, Compliance, and Talent Development That leads me to the other 2 roles we're hiring for, the dynamic duo of client delivery in our business. Next up: AI Strategist. You are the voice of the customer. If a customer success savant & world-class PM had a baby, it'd look like our AI Strategist. The job: Translate client goals into reality by uncovering bottlenecks and inefficiencies, aligning stakeholders, and executing change‑management and software‑implementation plans that unlock AI‑driven leverage. Must haves: deeply technical + previous experience as a technical consultant, PM, sales engineer, or similar role. Last up, is their technical partner: the Forward Deployed Engineer. Must-haves: elite engineering chops, AI-first mindset, relentless curiosity, team multiplier, client-facing. Oh...one more thing...your comp range is $250,000-$800,000. Why? It's uncapped (based on storypoint output). We pay you like a salesperson. Links to all jobs here: https://lnkd.in/edXZjyc8
-
Huy Khieu reacted on thisHuy Khieu reacted on thisUnpopular opinion, I'm not a fan of salary transparency regulation. Not even a little bit. 🤷 I can tell you exactly what will happen here. As soon as this comes in, aggregators will start to pick up the salary for every job ever advertised, turn those db's into a product, and sell them back to recruitment software vendors. In 18 months time, every time you apply for a job, the recruiter will get a little note from an AI that says "Jim is a Software Engineer level 2 at acme corp ltd, he likely earns £65,000" This kills your negotiating position when changing jobs dead in the water, and will ultimately mean slower wage growth as your career develops. IMO - its a trend that's been pushed by well meaning folks who are trying to drive equity and fair practise, but who don't realise how much it will bite applicants on the bum.
Experience
-
-
-
-
-
-
-
-
-
Education
-
University of Economics and Law
-
-
With the enjoyable between business and information system, my goal is to study for Master in MIS in the next 3 - 5 years
Current GPA: 7.5/10
Licenses & Certifications
Honors & Awards
-
2nd place in BI 2019 competition (sponsor by Shopee)
ITB
- Manage a team with 3 members
- Analyze business data, then define and present the solution to solve e-commerce business problem -
Certificate of joining P2A journey 2018
Multimedia University
Languages
-
English
-
Organizations
-
AIESEC
Organization Committee President Marcom
-
Other similar profiles
Explore more posts
-
OWA AI (formerly OtherWayAround)
𝗔/𝗕 𝘁𝗲𝘀𝘁𝗶𝗻𝗴 𝗶𝗻 𝘁𝗵𝗲 𝗔𝗽𝗽 𝗦𝘁𝗼𝗿𝗲 𝗶𝘀 𝗼𝗻𝗲 𝗼𝗳 𝘁𝗵𝗲 𝗺𝗼𝘀𝘁 𝗶𝗺𝗽𝗼𝗿𝘁𝗮𝗻𝘁 𝗴𝗿𝗼𝘄𝘁𝗵 𝗹𝗲𝘃𝗲𝗿𝘀, 𝗯𝘂𝘁 𝗼𝗻𝗹𝘆 𝗶𝗳 𝗶𝘁’𝘀 𝗱𝗼𝗻𝗲 𝗿𝗶𝗴𝗵𝘁. We just published a full breakdown of how to structure App Store experiments that actually move the needle. Inside the post: ✅ How to write strong hypotheses (with real examples) ✅ What not to test and why ✅ Why MDE and confidence intervals matter ✅ iOS vs. Google Play testing rules ✅ What to do after a test ends Whether you're testing screenshots, icons, or messaging, structured testing turns random ideas into reliable growth. 🔗 𝗥𝗲𝗮𝗱 𝘁𝗵𝗲 𝗳𝘂𝗹𝗹 𝗯𝗹𝗼𝗴 𝗽𝗼𝘀𝘁 𝗵𝗲𝗿𝗲: http://bit.ly/4kYJ6OX #AppGrowth #ABTesting #ASO #MobileMarketing #GrowthStrategy #CreativeTesting #ProductPageOptimization #UA #MobileApps
-
Interactive Labs, Inc
Is "Vibe Coding" the end of the traditional PRD? 🧐 At Interactive Labs, we recently tried something that flipped the traditional product-engineering workflow on its head. Instead of static Figma files and 50-page spec docs, our partners at Offset built "vibe-coded" prototypes—live, clickable code that expressed intent and behavior directly. What is "vibe coding"? It’s a PRD written in code. It’s messy, it’s minimal, and it’s meant to speak to engineers instead of just describing things in words. We took those "vibes" and turned them into production-ready, secure, and maintainable software. The result? ✅ Zero translation errors. ✅ Faster alignment between PMs and Engineers. ✅ Creative energy preserved from day one to launch. Read about how we’re bridging the gap between inspiration and execution: 👉 https://lnkd.in/giv33hCx
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content