bugyaluwang (@bugyaluwang) header
bugyaluwang (@bugyaluwang) profile picture

Full Stack Engineer | Pre @Tencent @miHoYo Up to this point, one still looks forward to a beautiful future. https://xitter.cf/t.co/TTWCqXHHnT

Online World
Joined October 2024
Photo by bugyaluwang (@bugyaluwang)Photo by bugyaluwang (@bugyaluwang)Photo by bugyaluwang (@bugyaluwang)Photo by bugyaluwang (@bugyaluwang)Photo by bugyaluwang (@bugyaluwang)Photo by bugyaluwang (@bugyaluwang)Photo by bugyaluwang (@bugyaluwang)Photo by bugyaluwang (@bugyaluwang)Photo by bugyaluwang (@bugyaluwang)Photo by bugyaluwang (@bugyaluwang)
用 devin 做的。 把自己的工作日志收拾成了一间小屋。 这里有工程、AI、产品体验的手记,也有做过的 Node.js、Rust 和 Agent 项目,还有一些正在形成的想法。 bugyaluwang.vercel.app #前端开发 #AI #独立开发 #Nodejs #agent
3
1
21
8,130
Google 再次证明了我的眼光😎第三 家中我最看好的。虽然中间略有波折
2
6
386
更正 御三家
46
claude 做视频真的很强,强到夸张。 gpt 和 gemini 用代码做视频的能力就确实惨不忍睹了...还时不时偷偷调 imagen 或者 nano 😂
1
296
刚刚跟一家公司 CEO 聊了一下。 他们在做数字员工,也就是工作的 Agent,可以理解成初期,你会有一个下属。 这让我大为震撼,因为在我看来,这种东西会是 Google 或者字节的下一步,甚至百度早有听说已经在做这个东西了。 AI power workspace 这个切入方向很好 但是我总感觉会直面大厂竞争...
9
1
19
3,960
“甚至百度”用词不准,百度是战略先锋,但总做不好
637
我目前的倾向度还是期望继续在当前公司。 主要是做的东西很棒,和我的个人项目契合度高达 90%,硬件+AI,C 端经验,我感觉我有高概率能学习到不错的东西。 如果是金融量化的话,说是量化 但是做的并非策略研究,而是内部交易系统一些金融监控相关的东西。招我很可能是看中了我的性能优化经验,有一点陷阱岗位的意思。 但我还是愿意往后聊一聊 反正八字还没一撇呢
有点烦恼。 今天晚上的面试过了,国内 top3 的金融量化。剩下一面要线下,感觉得请假。 但我已经接了 offer 甚至入职了头疼🥲
3
15
2,769
我要冷死了,这个温度跟股市正相关了是吗
414
希望有金融量化的佬指导一下我,对比互联网怎么样
有点烦恼。 今天晚上的面试过了,国内 top3 的金融量化。剩下一面要线下,感觉得请假。 但我已经接了 offer 甚至入职了头疼🥲
9
1
27
6,488
除了技术持续领先之后,大模型的用户没有忠诚度,复杂任务谁家的效果好就用谁家的,普通任务谁家的性价比高就用谁家的。 @grok 你觉得 claude 家目前拿下了我这里说的哪几个
2
696
@grok 那你觉得你拿下了哪一个
1
456
有点烦恼。 今天晚上的面试过了,国内 top3 的金融量化。剩下一面要线下,感觉得请假。 但我已经接了 offer 甚至入职了头疼🥲
27
1
135
30,887
想拒绝,又觉得是个好机会
2
9
3,830
前几天入职完毕。 之前有两周,连个面试都没有。而这两天拿到 offer 我主动中断流程之后,突然之间,有两家公司使劲约我面试。 😂一家说我笔试成绩好尚且有点原因,另一家 CEO 跟我约了线下沟通☺️双方都积极到我推辞不了 虽然不太恰当,但确实诠释了什么叫妻不如妾,妾不如偷,偷不如偷不着。
2
33
3,337
有时候我也很好奇,Senior 和 Junior 到底差在哪里?年限?or anything? 新的工作应该是 Senior 岗,但入职还是 mentor 带着,前两天也还没干活,就是纯看架构、看文档+体验产品。 我确实看困了
2
12
2,393
股灾了吗
3
674
我的观点:尽可能去大厂实习,多段最好。尤其对于院校并不占优势的同学。 这时候进一步衍生出如何去大厂实习?这一步可以通过去探索实际的实践经验来解决: 1. 比如去校内组织或者做一个项目去比赛,不要混。 2. 去 github 做开源,不要刷文档,多体验体验产品,很自然而然就会发现缺陷,然后可以写 PR,也可以去做自己喜欢的项目。 不过我这都是没有 AI 的时代的东西🤣
Replying to @bugyaluwang
第一份工作怎么找到好的
6
3
109
10,654
校招同学们一定要重视第一份工作,这决定了你的起点。 起点高,后边就算持续走下位替代,都是很不错的,你有自己选择的空间。 这个话我也曾经很苦口婆心地教给几个学弟😂但是他们那会儿真的是摆烂,现在工作一年了,好几个都在说后悔。 真的不要等到事教人,一遍就会,一遍就悔。
我感觉拒绝掉了一个非常好的机会...但大家各有看法吧。 大家只说 pdd 卷,没有说 pdd 的调薪,真的是半年能 +5k,一年固调一次,有机会调两次。 这种普调在其他公司就是纯梦😂
24
6
3
157
28,880
还没有和新同事深度合作,但是我蛮喜欢我的新同事的☺️比之前某群家伙真是好一百倍,希望之后也这样
2
3
1,208
Su obviously wants to play in the UK server, so I've decided to join them there too
Hi Bojie, thanks for the encouragement! Now that you’ve said it, I feel like I have to build it 😂 It may take a little while, though. My former internship mentor, who is also a good friend, helped me come up with the name OfferHunter. github.com/LofiSu/OfferHunte… 😂 I’ve been busy with other things lately, so this may not be released until December. If there’s enough demand, I’ll keep iterating and improving it. The idea is fairly straightforward: use the STAR method—Situation, Task, Action, Result—to explain your actual work clearly. Describe the context, what you were responsible for, what you specifically did, and the outcome, with evidence where available. Then use a job model—the target role’s responsibilities and competency requirements—to connect those experiences with your strengths and current circumstances. These are things my mentor taught me. He’s been a wonderful teacher and a very good friend. I think some online career advice makes the process sound more mysterious than it needs to be. If this would be useful, please give me a few days to put the approach together ☺️ In the meantime, I built recruiting-skills github.com/LofiSu/recruiting… for an HR professional with many years of experience at major tech companies, who now leads recruiting at a leading embodied AI company. It’s a separate, recruiter-facing toolkit for Claude Code, Codex, and Cursor. Given an arXiv paper, it researches the authors’ public professional profiles and produces an evidence-backed report. It can also filter an existing candidate index by graduation year, degree, and publication criteria. To try it, follow the README to install the skills, start a new agent session, and ask: “Research the authors of this arXiv paper and generate an HTML report.” Batch filtering requires preparing or importing a candidate index first; the repository doesn’t include a live candidate database. The outputs are for human review, not automated hiring decisions or outreach. I hope it’s helpful to others too! Since the requirements came from his workflow, it may not suit every recruiting scenario. Suggestions, feature requests, and feedback are very welcome by email or in the comments. And if you’d like to collaborate on an open-source project, please feel free to reach out anytime.
2
7
1,625