Justin Force

See the fastest LLM serving engine, 100% open source!!! If you’re running multi-round QA, RAG, or any long-context workload, you know the pain: slow time-to-first-token and expensive bills. You should take a look LMCache, a next-gen LLM serving engine extension designed to slash TTFT (time to first token) by 3–10x and boost throughput. How? LMCache stores KV caches from any reusable text (not just prefixes) across GPU, CPU DRAM, and disk, then shares them across serving instances. That means fewer recomputations, faster responses, and huge GPU savings. Key highlights: 🔥 Seamless integration with vLLM v1 ⚡ High-performance CPU KV offloading 🔗 P2P KVCache sharing across servers 💾 Storage on CPU, Disk, or NIXL ✅ Supported in vLLM production stack, llm-d, and KServe 📦 One-liner install: pip install lmcache If you’re scaling LLM inference, you should check it! 👉 GitHub link in the comments. #LLM #AI #Inference #OpenSource #RAG #AgenticAI

379

27 Comments

Hot take (with love): your To-Do app won’t get you hired. I’ve built them too. They’re great for learning CRUD, routes, and state. But the market’s seen a thousand. Hiring teams skim for signal. What they want to see: • You solved a real problem (yours or someone else’s) • End-to-end thinking (API design, auth, validation, tests) • You shipped (deploy + README + screenshots + a 60-sec demo) • You made it visible (why you built it, what’s next) It doesn’t have to be flashy—just useful and yours. How I’m doing it: FinLume. Not perfect. Not fancy. Still needs alot of work, But it solves my real headache (bills/subscriptions) and forced me into FastAPI, SQLAlchemy, auth, and testing. That’s the point. Simple non-To-Do ideas you could ship: • Renovation expense tracker with monthly summary (been living this! 🧱) • Job/quote scheduler for trades (my gardening days inspired this) • Interview tracker with follow-up email templates • Subscription cancel-reminder + renewal timeline Bottom line: Different > generic. Visible > invisible. Post progress. Ask for feedback. Make noise without screaming. #JuniorDeveloper #BuildInPublic #PortfolioProjects #LearnInPublic #PythonDeveloper #FastAPI

114

79 Comments

The biggest lie in tech hiring: "We hire the best developers." No, you don't. You hire the developers who interview the best. I've worked with brilliant developers who can build entire systems from scratch but freeze up in technical interviews. They know every design pattern. They've solved complex production issues. They've mentored entire teams. But they can't explain FizzBuzz under pressure. Meanwhile, less experienced developers who practiced mock interviews land the $150k+ offers. Here's what I discovered after helping 100+ developers: The interview is a completely different skill than actual development. And most experienced developers never learn it. They think their GitHub speaks for itself. They assume their experience is enough. They walk into interviews unprepared. Then they wonder why they're getting lowball offers or straight rejections. The solution isn't becoming a better developer. It's becoming a better interviewer. After working across different continents with developers at every level, I've seen the patterns. The ones who succeed treat interviews like any other skill: They practice. They prepare. They get feedback. They improve. Want my complete interview preparation framework that helped developers increase their offers by $50k-$80k? Comment "INTERVIEW" and I'll send you the free guide. Stop letting poor interview skills cap your career.

96

171 Comments

You may see folks claiming that they can get "A.I." coding assistants to do exactly what they want by completely and precisely describing the software they require up-front, and hailing the return of Big Design Up-Front. Anyone who's spent more than a handful of hours using these tools will know that this simply isn't true.

103

116 Comments

Here are 5 NEW companies paying $100k-$400k for FULLY-REMOTE devs: 1. Circle 2. The Linux Foundation 3. Form3 4. Assured 5. Qualio They get it. - They don't have the FAANG brand. - They even pay less than FAANGs in the US. Yet, they're able to easily hire and retain FAANG-level talent. Want to know about other companies in this category? EuroTopTech.com lists thousand jobs from 100+ such companies. It also tracks thousand top-paying onsite, hybrid and country-remote jobs. In Europe. It also provides data regarding compensation and saving rates. Check it out if it sounds interesting 🙂 ♻️ Like/Comment/Share if you want to help people in your network.

232

17 Comments

styled-components has entered maintenance mode. Its maintainer was refreshingly honest: "Don't use this for new projects." Respect. But what about the millions of components already in production? My colleague Cody recognized the problem immediately. He submitted a PR with React 18 optimizations back in July 2024. With maintenance mode, that PR wasn't going anywhere. So Cody did what great engineers do. He turned our lingering PR into a fork, optimized it further, then opened it up for everyone. Linear tested it: 40% faster initial renders. No big code changes needed. The kicker? styled-components never implemented React 18's useInsertionEffect - a hook specifically designed for CSS-in-JS performance. It shipped in March 2022. Cody PR'd it fixed it in a day. We're calling it styled-components-last-resort because that's what it is. Not a new maintenance path. Not a forever solution. Just a performance bridge while teams migrate properly. Sometimes open source isn't about building the future. Sometimes it's about helping each other manage the present. Massive credit to Cody Olsen for the implementation, Kenneth Skovhus at Linear for testing, and Evan Jacobs for maintaining styled-components for years. Technical deep dive: https://lnkd.in/dk4CfsP8 GitHub: https://lnkd.in/duzY-Vy6 What's your styled-components migration plan?

111

3 Comments