[{"data":1,"prerenderedAt":345},["ShallowReactive",2],{"project-43296918":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":36,"aspects":37,"strip":43,"post_text":46,"what":47,"said":75,"maker_in_thread":221,"shape":229,"grid":233,"attempts":307,"same_week":308,"maker_shows":32,"maker_hidden":337,"reader_version":338,"card":339},"43296918","rlama","Open-Source DocumentAI with Ollama","title","Show HN: Open-Source DocumentAI with Ollama","https:\u002F\u002Frlama.dev\u002F","rlama.dev","Dontizi",1741399933,2025,295,33,188,3837,5,"legend","ai","retrieval augmented generation","cli",[24,25,26,27,28],"rag","ollama","local llm","document qa","vector search","curious",-0.135,{"positive":32,"neutral":33,"mixed":34,"critical":35},2,12,0,6,20,{"execution":38,"idea":39,"performance":40,"privacy":41,"docs":42},{"positive":34,"critical":32},{"positive":32,"critical":32},{"positive":34,"critical":32},{"positive":34,"critical":34},{"positive":34,"critical":34},[44,45,45,34,34,45,34,34,45,45,45,34,44,34,44,34,34,44,34,18,34,45,34,45,45,34,44,18,45,44,45,45,34],1,3,null,{"problem":46,"audience":46,"pricing":46,"motivation":46,"time_to_build":46,"stack":48,"openness":72,"stage":73,"team_size":74},[49,56,60,64,68],{"value":50,"source":51,"evidence":52},"Go","maker",[53],{"id":54,"by":11,"text":55},"43298152","Hey! Yes, that's something I was planning to do—a complete documentation on the code, its architecture, and the entire stack to allow others to develop alongside me. I just…",{"value":57,"source":51,"evidence":58},"Cobra CLI framework",[59],{"id":54,"by":11,"text":55},{"value":61,"source":51,"evidence":62},"Ollama API for embeddings and completions",[63],{"id":54,"by":11,"text":55},{"value":65,"source":51,"evidence":66},"JSON files on local filesystem for storage",[67],{"id":54,"by":11,"text":55},{"value":69,"source":51,"evidence":70},"Custom cosine similarity vector search",[71],{"id":54,"by":11,"text":55},"open_source","beta","unknown",{"praise":76,"objections":100,"requests":132,"questions":166,"alternatives":186},[77,89],{"value":78,"source":79,"evidence":80},"Single binary distribution and Go implementation appreciated","commenter",[81,85],{"id":82,"by":83,"text":84},"43298080","foundzen","What is the architecture\u002Ftech-stack used in building this? I didn't find this info neither on github readme, nor on website. I like the fact that it is written in Go and small…",{"id":86,"by":87,"text":88},"43298997","mentalgear","Really nice project, congrats & great work! Quick notes: - as an end user, some primary concerns re apps using the file system: ``` - who will be able to read it? does the app…",{"value":90,"source":79,"evidence":91},"Useful idea of local document AI",[92,96],{"id":93,"by":94,"text":95},"43298273","TheNicholasNick","This is cool and pretty much what I was wondering about, I mean obvious it was possible, cool to see it implemented. Looking forward to having a play. I am building a tool purely…",{"id":97,"by":98,"text":99},"43297885","ding444","This is great! It would be great if there were an API interface for integration into other systems.",[101,108,115,125],{"value":102,"source":79,"evidence":103},"No chunking; whole documents embedded and truncated, so retrieval fails on long books",[104],{"id":105,"by":106,"text":107},"43300305","andai","This appears to do no chunking. It just shoves the entire document (entire book, in my case) into the embedding request to Ollama. So it's only helpful if all your documents are…",{"value":109,"source":79,"evidence":110},"Hallucination risk and lack of source citation",[111],{"id":112,"by":113,"text":114},"43299285","smusamashah","I feel very doubtful on usefulness of these tools because of hallucinations. How reliable is this one in comparison with others like these? How well does it cite the source? To…",{"value":116,"source":79,"evidence":117},"Requires Python dependencies for text extraction, undermining single binary",[118,121],{"id":119,"by":106,"text":120},"43300074","rlama requires a python install (and several dependencies via pip) to extract text. https:\u002F\u002Fgithub.com\u002FDonTizi\u002Frlama\u002Fblob\u002Fmain\u002Finternal\u002Fservice\u002Fdocument_loader.go#L448",{"id":122,"by":123,"text":124},"43298426","Xiol32","Defeats the point of the single binary installation if you have to set up dependencies.",{"value":126,"source":79,"evidence":127},"Custom vector store less performant than established options",[128],{"id":129,"by":130,"text":131},"43298265","dcreater","Why not use an established open source vector db like pg_vector etc? I imagine your implementation is not going to be as performant",[133,140,150,157,162],{"value":134,"source":79,"evidence":135},"Add chunking with overlap",[136,137],{"id":105,"by":106,"text":107},{"id":138,"by":11,"text":139},"43306194","I have just implemented chunking with overlap for larger documents to split texts into smaller chunks and ensure access to all documentation in your RAG. It's currently in the…",{"value":141,"source":79,"evidence":142},"Support llama.cpp backend",[143,147],{"id":144,"by":145,"text":146},"43298104","car","Could this work with llama.cpp, since it’s the engine behind Ollama? I usually build llama.cpp from source and download quantized (GGUF) models from Huggingface, haven’t used…",{"id":148,"by":145,"text":149},"43298435","That would be great. Llama.cpp’s built in server offers HTTP embedding endpoints.",{"value":151,"source":79,"evidence":152},"Add an API interface",[153,154],{"id":97,"by":98,"text":99},{"id":155,"by":11,"text":156},"43297902","I thought about adding an API interface for it. It is on my to-do list of things that could be good to add. For now, I'm gathering feedback to see what people like about it or not.",{"value":158,"source":79,"evidence":159},"Show search results and cite sources",[160,161],{"id":105,"by":106,"text":107},{"id":112,"by":113,"text":114},{"value":163,"source":79,"evidence":164},"Support TypeScript files and privacy\u002Ffile access guarantees",[165],{"id":86,"by":87,"text":88},[167,171,178],{"value":168,"source":79,"evidence":169},"What is the architecture and tech stack",[170],{"id":82,"by":83,"text":84},{"value":172,"source":79,"evidence":173},"What license is it released under",[174],{"id":175,"by":176,"text":177},"43297435","mattnewton","Cool project. What license is this released under? Not seeing it documented.",{"value":179,"source":79,"evidence":180},"How good and reliable is the RAG answer quality",[181,185],{"id":182,"by":183,"text":184},"43300263","KaoruAoiShiho","How good is the RAG though? Just throwing a vector db at it doesn't make it useful...",{"id":112,"by":113,"text":114},[187,194,201,207,214],{"value":188,"source":79,"evidence":189},"Open WebUI document database",[190],{"id":191,"by":192,"text":193},"43299504","tarruda","> How well does it cite the source? I don't know about the OP tool, but open webui has its own document database which you can integrate with LLMs, and when answering questions…",{"value":195,"source":79,"evidence":196},"AWS Bedrock knowledge bases",[197],{"id":198,"by":199,"text":200},"43301536","danihh","Not a paid solution, but great for testing models yourself: AWS bedrock. Wonky documentation (definitely released too early), but imo the best model agnostic diy solution out…",{"value":202,"source":79,"evidence":203},"Orama hybrid search",[204],{"id":205,"by":87,"text":206},"43298844","I recommend using this hybrid vector\u002Ffull text search engine that works across many runtimes: https:\u002F\u002Fgithub.com\u002Foramasearch\u002Forama",{"value":208,"source":79,"evidence":209},"chromem-go embedded vector store",[210],{"id":211,"by":212,"text":213},"43299005","PhilippGille","Hi, if you want to keep using a Go embedded\u002Fin-process vector store, but with some additional features, you can check out my project https:\u002F\u002Fgithub.com\u002Fphilippgille\u002Fchromem-go",{"value":215,"source":79,"evidence":216},"Building your own RAG with Ollama tutorial",[217],{"id":218,"by":219,"text":220},"43303993","magic_hamster","Just do your own RAG. It's very easy and Ollama actually have a quick start tutorial on their page. Then you can also fine-tune the process to your needs.",{"comments":222,"answered":45,"key":223},7,[224,227,228],{"id":225,"text":226},"43305240","This is my next step. Currently, I’ve built an MVP to test the features, integrations, and see how far I can go with rLlama. I’m already developing a RAG on my end by chunking…",{"id":138,"text":139},{"id":54,"text":55},{"commenters":230,"max_depth":18,"maker_replies":222,"first_comment_mins":231,"lifespan_hours":232},24,116,26,[234,236,240,244,248,249,253,255,256,258,261,263,264,266,268,269,270,271,273,274,277,279,280,282,283,285,287,288,291,295,297,301,305],[105,34,44,235,106,34],"project\u002Fcritical\u002Fexecution",[237,34,45,238,239,34],"43301000","project\u002Fneutral\u002Fidea","rafaelmn",[241,34,45,242,243,34],"43301051","project\u002Fneutral\u002Fperformance","simonw",[245,34,34,246,247,34],"43301233","other\u002Fneutral\u002Fnone","elliot07",[198,34,34,246,199,34],[250,34,45,251,252,34],"43304395","project\u002Fneutral\u002Fexecution","codelion",[225,34,34,254,11,44],"maker\u002Fneutral\u002Fexecution",[138,34,34,254,11,44],[86,44,45,257,87,34],"project\u002Fneutral\u002Fprivacy",[259,44,45,257,260,34],"43300996","ImPostingOnHN",[82,32,45,262,83,34],"maker\u002Fneutral\u002Fdocs",[54,32,34,251,11,44],[129,32,44,265,130,34],"project\u002Fcritical\u002Fperformance",[122,32,34,267,123,34],"commenter\u002Fcritical\u002Fexecution",[119,32,44,235,106,34],[205,32,34,246,87,34],[211,32,34,246,212,34],[112,45,44,272,113,34],"project\u002Fcritical\u002Fidea",[191,45,34,246,192,34],[97,275,18,276,98,34],4,"maker\u002Fpositive\u002Fidea",[155,275,34,278,11,44],"maker\u002Fneutral\u002Fidea",[144,18,45,238,145,34],[281,18,34,278,11,44],"43298132",[148,18,45,238,145,34],[175,35,45,284,176,34],"maker\u002Fneutral\u002Fnone",[286,35,34,284,11,44],"43297836",[218,222,44,272,219,34],[93,289,18,290,94,34],8,"project\u002Fpositive\u002Fidea",[292,293,45,238,294,34],"43302039",9,"jwithington",[182,296,44,265,183,34],10,[298,299,45,238,300,34],"43298918",11,"lclc",[302,33,45,303,304,34],"43298520","project\u002Fneutral\u002Fdocs","sorokod",[306,33,34,262,11,44],"43298633",[],[309,313,318,323,328,332],{"id":310,"name":311,"points":312,"category":20},"43331582","Factorio Learning Environment",749,{"id":314,"name":315,"points":316,"category":317},"43328095","Seven39",735,"consumer_social",{"id":319,"name":320,"points":321,"category":322},"43347306","Time Portal",497,"games",{"id":324,"name":325,"points":326,"category":327},"43280615","CodeTracer",334,"dev_tools",{"id":329,"name":330,"points":331,"category":317},"43304785","daale.club",302,{"id":333,"name":334,"points":335,"category":336},"43333661","Pila",301,"hardware",false,"v2",{"id":340,"name":5,"line":6,"ai":337,"cat":20,"year":13,"points":14,"comments":15,"rec":29,"s":341,"top":18,"kw":342,"maker":11,"domain":10,"tier":19,"attempts":34,"strip":343,"url":9,"mix":344,"praise":78,"objection":102,"sub":21},43296918,-0.13,[24,25,26,27,28],"133003003330101001050303301531330",[32,33,34,35],1793991114609]