[{"data":1,"prerenderedAt":338},["ShallowReactive",2],{"project-37034575":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":10,"maker":11,"posted_at":12,"year":13,"points":14,"comments":15,"rank_in_year":16,"shown_in_year":17,"top":18,"tier":19,"category":20,"subcategory":21,"format":22,"keywords":23,"reception":29,"sentiment":30,"stance_mix":31,"scored_comments":34,"aspects":35,"strip":42,"post_text":44,"what":45,"said":80,"maker_in_thread":200,"shape":210,"grid":214,"attempts":301,"same_week":302,"maker_shows":37,"maker_hidden":331,"reader_version":332,"card":333},"37034575","Axilla","Open-source TypeScript framework for LLM apps","title","Show HN: Axilla – Open-source TypeScript framework for LLM apps","https:\u002F\u002Fgithub.com\u002Faxilla-io\u002Fax","github.com","nichochar",1691416838,2023,161,40,377,3109,13,"notable","ai","llm application framework","library",[24,25,26,27,28],"typescript llm framework","rag","llm evaluation","vector store","openai typescript","debated",0,{"positive":32,"neutral":33,"mixed":30,"critical":33},6,7,20,{"idea":36,"docs":39,"performance":40},{"positive":37,"critical":38},4,5,{"positive":30,"critical":30},{"positive":30,"critical":41},1,[38,30,30,38,30,38,30,43,30,30,41,43,30,30,43,41,30,30,41,41,43,43,43,41,41,30,30,38,41,38,38,43,30,30,30,30,30,30,30,30],3,"Hi HN, we are Nick and Ben, creators of Axilla - an open source TypeScript framework to develop LLM applications. It’s in the early stages but you can use it today: we’ve already published 2 modules and have more coming soon.\n\nBen and I met while working at Cruise on the ML platform for self-driving cars. We spent many years there and learned the hard way that shipping AI is not quite the same as shipping regular code. There are many parts of the ML lifecycle, e.g., mining, processing, and labeling data and training, evaluating, and deploying models. Although none of them are rocket science, m",{"problem":46,"audience":52,"pricing":59,"motivation":60,"time_to_build":59,"stack":64,"openness":77,"stage":78,"team_size":79},{"value":47,"source":48,"evidence":49},"Shipping LLM applications efficiently across the ML lifecycle, from data retrieval to evaluation","maker",[50],{"id":51},"post",{"value":53,"source":48,"evidence":54},"Software teams building LLM applications in TypeScript",[55,56],{"id":51},{"id":57,"by":11,"text":58},"37035833","We're very open to contributions, I am interested in what the integration would look like. Do you want to email me at nicholas@axilla.io? we can get into the details.",null,{"value":61,"source":48,"evidence":62},"Teams building LLM apps in TypeScript lack good tooling, and the makers want an end-to-end platform based on their Cruise ML experience",[63],{"id":51},[65,69],{"value":66,"source":48,"evidence":67},"TypeScript",[68],{"id":51},{"value":70,"source":71,"evidence":72},"Node.js","commenter",[73],{"id":74,"by":75,"text":76},"37042169","benjreinhart","Ahh yes, right now we're operating at a higher-level of the stack. That said, we are investigating serving from Node and possibly on edge devices with WebGPU. For serving from…","open_source","beta","team",{"praise":81,"objections":108,"requests":155,"questions":164,"alternatives":176},[82,97],{"value":83,"source":71,"evidence":84},"TypeScript-first approach is welcome",[85,89,93],{"id":86,"by":87,"text":88},"37035318","mlejva","Hey Nick and Ben, congrats to launch! I really like that you're going in the TS way by default. I personally think there will me more AI Engineers (devs building LLM apps\u002Fagents)…",{"id":90,"by":91,"text":92},"37037388","jumploops","We use GPT-4 pretty heavily in a Typescript project, but have noticed lag from the TS versions of popular libraries (OpenAI’s npm lib, Langchain TS, etc.). This framework is…",{"id":94,"by":95,"text":96},"37046943","fosterfriends","Congrats on building the library! I’ve recently been playing around with the js implementation of langchain, and I’m excited for there to be more high quality typescript support…",{"value":98,"source":71,"evidence":99},"Demo looks promising",[100,104],{"id":101,"by":102,"text":103},"37035260","rzmmm","Hi, I checked out the demo and it looks very promising. As someone who is not very familiar with AI development, I feel a bit puzzled looking at the code examples. If I use the…",{"id":105,"by":106,"text":107},"37036892","astrodude","Hi Nick and Ben, Looked at the demo. Great job! I'll be trying it soon!",[109,128,135,145],{"value":110,"source":71,"evidence":111},"Name means armpit in Spanish and Portuguese",[112,116,120,124],{"id":113,"by":114,"text":115},"37041653","kwanbix","What a horrible name. Why? Just why? You are putting so much love into something to give it a name that is so bad?",{"id":117,"by":118,"text":119},"37044449","dietr1ch","Yeah, I'd reconsider this too. If this sounds too opinionated so far, I'll give you a fact. As a Spanish speaker, this is also a little too close to the word for armpit, and I'd…",{"id":121,"by":122,"text":123},"37047706","capitanazo77","Feedback: Axila in Spanish means armpit",{"id":125,"by":126,"text":127},"37034830","SwiftyBug","\"Axilla\" in Portuguese means \"armpit\" (it's spelt \"axila\"). I like the name more because of this. Congrats on the launch! As a developer who's been working a lot with TypeScript…",{"value":129,"source":71,"evidence":130},"Abstraction layers over LLM APIs add little value",[131],{"id":132,"by":133,"text":134},"37037541","fswd","I've come to the conclusion that anything that \"abstracts\" the openai complete\u002Fchat complete API call is just bad practice and to stay away from the entire framework, with the…",{"value":136,"source":71,"evidence":137},"Founders' Cruise background is weak credential for RAG work and the pitch reads as hype",[138,142],{"id":139,"by":140,"text":141},"37042354","BoorishBears","My two cents which you are free to ignore (and I almost implore you to ignore): I'm sure this is useful but as someone who works at an AV company and is building with these new…",{"id":143,"by":140,"text":144},"37043834","I just don't see many people shipping AI software right now in the way AI applies to AVs. But I'm also giving you my two cents as someone who's not only building in the space,…",{"value":146,"source":71,"evidence":147},"Performance of JavaScript for ML tasks is questioned",[148,152],{"id":149,"by":150,"text":151},"37039345","fullstackchris","I'm gonna be that guy who will probably show up sooner or later anyway, but... I can't imagine performance can compete with other languages? What were your findings or experience…",{"id":153,"by":150,"text":154},"37041417","I'm no expert in actual ML implementations but I was under the impression that Python (i.e Tensorflow) is actually C\u002FC++ based under the hood. I just meant I can't imagine the V8…",[156,160],{"value":157,"source":71,"evidence":158},"Caching layer for deterministic responses",[159],{"id":90,"by":91,"text":92},{"value":161,"source":71,"evidence":162},"Integrations via PRs",[163],{"id":86,"by":87,"text":88},[165,169],{"value":166,"source":71,"evidence":167},"Can LLM outputs be trusted to be well structured from templated prompts?",[168],{"id":101,"by":102,"text":103},{"value":170,"source":71,"evidence":171},"Why is langchain-js considered bloatware?",[172],{"id":173,"by":174,"text":175},"37036392","zarazas","Why do you think langchainjs is bloatware?",[177,189,193],{"value":178,"source":71,"evidence":179},"LangChain",[180,184,185],{"id":181,"by":182,"text":183},"37035135","savy91","Amazing! I am working on projexts where we use LLMs and Typescript and and besides langchain-js which can be only described as bloatware, I can never find anything and find…",{"id":90,"by":91,"text":92},{"id":186,"by":187,"text":188},"37038906","jawerty","Yea learning how to use the core API directly should be the focus for any engineer. Lots of frameworks built on top of LLMs are being made very quickly each with their own…",{"value":190,"source":71,"evidence":191},"Microsoft Guidance",[192],{"id":132,"by":133,"text":134},{"value":194,"source":71,"evidence":195},"Calling the OpenAI API directly with fetch",[196,199],{"id":197,"by":91,"text":198},"37039286","Very cool. To be completely candid, we just hit OpenAI directly, no 3rd party libs involved at the moment (just fetch). We're open to trying more TS-focused libraries, but…",{"id":132,"by":133,"text":134},{"comments":43,"answered":201,"key":202},2,[203,204,207],{"id":57,"text":58},{"id":205,"text":206},"37038618","Glad to hear this, indeed we think there's opportunity for some more cutting edge tooling in the TS ecosystem. We absolutely want to add a caching layer. Actually, we think…",{"id":208,"text":209},"37043557","Thanks for the feedback! The lesson that we learned at Cruise is that the tough thing when shipping AI software is closing the data loop and integrating all of the steps of the…",{"commenters":211,"max_depth":38,"maker_replies":43,"first_comment_mins":212,"lifespan_hours":213},30,21,22,[215,217,219,222,226,228,229,230,232,234,235,237,239,240,243,245,247,250,251,252,254,257,258,259,260,261,264,267,269,271,273,275,278,281,284,287,289,292,295,298],[86,30,38,216,87,30],"project\u002Fpositive\u002Fidea",[57,30,30,218,11,41],"maker\u002Fneutral\u002Fnone",[220,30,30,221,87,30],"37035893","commenter\u002Fneutral\u002Fnone",[223,30,38,224,225,30],"37040627","project\u002Fpositive\u002Fnone","williamzeng0",[227,30,30,221,87,30],"37044735",[90,41,38,216,91,30],[205,41,30,216,11,41],[197,41,43,231,91,30],"project\u002Fneutral\u002Fidea",[181,201,30,233,182,30],"other\u002Fneutral\u002Fnone",[173,201,30,221,174,30],[132,43,41,236,133,30],"project\u002Fcritical\u002Fidea",[238,43,43,231,75,30],"37038485",[186,43,30,233,187,30],[241,43,30,221,242,30],"37043681","victor106",[101,37,43,244,102,30],"project\u002Fneutral\u002Fdocs",[139,38,41,246,140,30],"maker\u002Fcritical\u002Fnone",[248,38,30,221,249,30],"37043096","iamnafets",[208,38,30,221,11,41],[143,38,41,236,140,30],[149,32,41,253,150,30],"project\u002Fcritical\u002Fperformance",[255,32,43,256,75,30],"37039869","project\u002Fneutral\u002Fperformance",[153,32,43,256,150,30],[74,32,43,256,75,30],[113,33,41,236,114,30],[117,33,41,236,118,30],[262,33,30,221,263,30],"37044598","mattigames",[265,33,30,233,266,30],"37046722","cafard",[94,268,38,216,95,30],8,[121,270,41,236,122,30],9,[105,272,38,224,106,30],10,[125,274,38,216,126,30],11,[276,274,43,231,277,30],"37034927","giovannylira",[279,274,30,233,280,30],"37035199","whstl",[282,274,30,233,283,30],"37035350","mattsan",[285,274,30,233,286,30],"37035957","anthk",[288,274,30,233,126,30],"37037579",[290,274,30,233,291,30],"37035570","whatrocks",[293,274,30,233,294,30],"37038186","jw1224",[296,274,30,233,297,30],"37035640","wbkang",[299,274,30,233,300,30],"37037289","carbocation",[],[303,308,313,317,322,327],{"id":304,"name":305,"points":306,"category":307},"36987920","Hydra 1.0",269,"data",{"id":309,"name":310,"points":311,"category":312},"37034951","Phind V2",221,"dev_tools",{"id":314,"name":315,"points":316,"category":20},"37050532","Airbyte",220,{"id":318,"name":319,"points":320,"category":321},"37064152","ToSNotify",190,"productivity",{"id":323,"name":324,"points":325,"category":326},"37062422","Putting the \"You\" in CPU",187,"learning_science",{"id":328,"name":329,"points":330,"category":20},"37038053","ChainForge",177,false,"v2",{"id":334,"name":5,"line":6,"ai":331,"cat":20,"year":13,"points":14,"comments":15,"rec":29,"s":30,"top":18,"kw":335,"maker":11,"domain":10,"tier":19,"attempts":30,"strip":336,"url":9,"mix":337,"praise":83,"objection":110,"sub":21},37034575,[24,25,26,27,28],"5005050300130031001133311005155300000000",[32,33,30,33],1793991071091]