[{"data":1,"prerenderedAt":293},["ShallowReactive",2],{"blog-post-en-barbershop-appointment-intervals":3,"blog-siblings-en-barbershop-appointment-intervals":96},{"id":4,"title":5,"author":6,"body":7,"category":78,"cover":79,"date":80,"description":81,"draft":82,"extension":83,"meta":84,"navigation":85,"path":86,"readingTime":87,"seo":88,"stem":89,"tags":90,"__hash__":95},"blog_en\u002Fen\u002Fblog\u002Fbarbershop-appointment-intervals.md","What we learned from 850 barbershops about appointment intervals","Codifya",{"type":8,"value":9,"toc":70},"minimark",[10,18,23,30,37,41,44,60,63,67],[11,12,13,14],"p",{},"We examined appointment data from over 850 barbershops using BerberBul. The question was simple: ",[15,16,17],"strong",{},"what is the most productive appointment interval?",[19,20,22],"h2",{"id":21},"findings","Findings",[11,24,25,26,29],{},"The highest average occupancy rate was among barbers working with ",[15,27,28],{},"35-minute intervals",". Shops using intervals under 30 minutes saw lower customer satisfaction — especially with services that include hair washing. Shops with intervals above 45 minutes had reduced daily capacity.",[11,31,32,33,36],{},"Unexpected finding: ",[15,34,35],{},"the day before Sunday"," (Saturday) was the busiest. But the most productive day was Tuesday. Tuesday's average occupancy was 82%, while Saturday was 67% — busier but less profitable because wait times stretched.",[19,38,40],{"id":39},"what-we-recommend","What we recommend",[11,42,43],{},"Two solutions for Saturday:",[45,46,47,54],"ol",{},[48,49,50,53],"li",{},[15,51,52],{},"Move earlier appointments to Saturday morning"," — 09:00-11:00 has the highest booking rate.",[48,55,56,59],{},[15,57,58],{},"Use time blocks"," — instead of 30-minute slots, use 45-minute \"peak period blocks.\"",[11,61,62],{},"Barber shops using BerberBul's automatic conflict detection with this kind of planning see 23% lower cancellation rates.",[19,64,66],{"id":65},"the-real-lesson-from-the-data","The real lesson from the data",[11,68,69],{},"Appointment system design varies by barbershop physical capacity and staff experience. There's no single right answer; but with the right data, each business can find its own. BerberBul makes this measurable.",{"title":71,"searchDepth":72,"depth":72,"links":73},"",3,[74,76,77],{"id":21,"depth":75,"text":22},2,{"id":39,"depth":75,"text":40},{"id":65,"depth":75,"text":66},"BerberBul",null,"2026-05-02","We analyzed appointment behavior from 850+ barbershops on BerberBul. The optimal interval turned out to be different from what most shops assumed.",false,"md",{},true,"\u002Fen\u002Fblog\u002Fbarbershop-appointment-intervals",5,{"title":5,"description":81},"en\u002Fblog\u002Fbarbershop-appointment-intervals",[91,92,93,94],"data","operations","scheduling","insights","ag4fwhZ8V1QMliIrvt7N4fp8E0Bt3ohDCPlqN42xjNY",{"prev":97,"next":209},{"id":98,"title":99,"author":6,"body":100,"category":195,"cover":79,"date":196,"description":197,"draft":82,"extension":83,"meta":198,"navigation":85,"path":199,"readingTime":200,"seo":201,"stem":202,"tags":203,"__hash__":208},"blog_en\u002Fen\u002Fblog\u002Fflowbit-mcp-ile-otomatik-gorev-yonetimi.md","FlowBit and MCP: automated task management that works",{"type":8,"value":101,"toc":189},[102,105,109,112,115,119,122,134,138,150,153,164,167,171,182],[11,103,104],{},"Shipping yet another project management tool meant answering two questions: where to put AI, and where not to. FlowBit is the product of those two decisions.",[19,106,108],{"id":107},"why-another-pm-tool","Why another PM tool",[11,110,111],{},"Most teams using Jira, Linear, or Asana kept saying the same thing: the tool can do too much, and the team does too little of it. Complexity slows small teams down and forces large teams to rewrite their own processes.",[11,113,114],{},"When designing FlowBit we prioritized: few but right features, and AI assistance exactly where users touch the product daily.",[19,116,118],{"id":117},"why-mcp-matters","Why MCP matters",[11,120,121],{},"MCP (Model Context Protocol) creates a standard bridge between AI models and applications:",[123,124,125,128,131],"ul",{},[48,126,127],{},"Claude, GPT-4, or a local model (Ollama) connect through one integration.",[48,129,130],{},"The model can see in-app context (team, sprint, ticket) without separate system prompts.",[48,132,133],{},"Data scoping is explicit — the model only reads what's allowed.",[19,135,137],{"id":136},"what-we-observed-in-the-field","What we observed in the field",[11,139,140,141,145,146,149],{},"After three months with pilot teams, the clearest finding: AI adds the most value not at the ",[142,143,144],"em",{},"deciding"," moment, but at the ",[142,147,148],{},"remembering"," one.",[11,151,152],{},"A typical flow:",[45,154,155,158,161],{},[48,156,157],{},"Meeting notes get transcribed.",[48,159,160],{},"FlowBit suggests likely tasks and owners from the text.",[48,162,163],{},"Sprint reports — who did what, where it stalled — drop into Slack automatically.",[11,165,166],{},"None of these steps is \"smart\" individually. Together, they erase the team's \"we forgot about that\" moments.",[19,168,170],{"id":169},"what-we-learned","What we learned",[123,172,173,176,179],{},[48,174,175],{},"Auto-suggested tasks must always pass through human approval. AI shouldn't create tickets alone.",[48,177,178],{},"Reports should be short. Anything past 8 lines doesn't get read.",[48,180,181],{},"Local mode (Ollama) is in demand — especially in the public and healthcare sectors.",[11,183,184,185,188],{},"FlowBit is still early. But positioning AI as a ",[142,186,187],{},"reminder"," rather than an accelerator is working.",{"title":71,"searchDepth":72,"depth":72,"links":190},[191,192,193,194],{"id":107,"depth":75,"text":108},{"id":117,"depth":75,"text":118},{"id":136,"depth":75,"text":137},{"id":169,"depth":75,"text":170},"FlowBit","2026-05-08","The useful part of AI-assisted task creation is making it easier for the team to decide where to focus. FlowBit's design choices and what we observed in the field.",{},"\u002Fen\u002Fblog\u002Fflowbit-mcp-ile-otomatik-gorev-yonetimi",6,{"title":99,"description":197},"en\u002Fblog\u002Fflowbit-mcp-ile-otomatik-gorev-yonetimi",[204,205,206,207],"ai","mcp","product management","automation","eEPaJPHkXnNQIk4MQTX7y9-HiSS6VIOKRQGV8lN51P0",{"id":210,"title":211,"author":6,"body":212,"category":280,"cover":79,"date":281,"description":282,"draft":82,"extension":83,"meta":283,"navigation":85,"path":284,"readingTime":285,"seo":286,"stem":287,"tags":288,"__hash__":292},"blog_en\u002Fen\u002Fblog\u002Fcityos-vatandas-bildirim-sistemi.md","Designing a citizen reporting system from scratch",{"type":8,"value":213,"toc":273},[214,217,220,224,227,231,234,241,245,248,255,259,262,266],[11,215,216],{},"Citizen reporting is one of the most visible service metrics for a municipality — and usually one of the weakest operationally. Call center, social media, mobile app: each channel tends to live in its own silo.",[11,218,219],{},"When designing CityOS's reporting module, we set out to solve three core problems.",[19,221,223],{"id":222},"_1-a-unified-queue","1. A unified queue",[11,225,226],{},"Which channel a citizen used — phone, mobile app, web — should be a detail the assignee doesn't have to think about. CityOS pipes every channel into one queue; the assigned report looks the same regardless of source.",[19,228,230],{"id":229},"_2-auto-classification-but-overridable","2. Auto-classification, but overridable",[11,232,233],{},"Whether a report is \"park maintenance\" or \"road maintenance\" is obvious in roughly 80% of cases. The remaining 20% needs a human.",[11,235,236,237,240],{},"CityOS auto-classifies the incoming report, but the assignee can always change the category. That small detail — ",[142,238,239],{},"\"AI suggested it, but you have the final say\""," — dramatically increases acceptance.",[19,242,244],{"id":243},"_3-live-updates-for-the-citizen","3. Live updates for the citizen",[11,246,247],{},"When the internal process completes, push notifications are the fastest way to tell the citizen. But which step should we notify? Assigned, in-progress, completed?",[11,249,250,251,254],{},"Three pilot municipalities tested it and the result was clear: a single notification — ",[142,252,253],{},"\"Your report is complete\""," — produced higher satisfaction than multiple updates. Notifying every step felt like nagging.",[19,256,258],{"id":257},"the-invisible-part-gdprkvkk","The invisible part: GDPR\u002FKVKK",[11,260,261],{},"Separating identity from report content, enabling audit log from day one, and minimizing personal data — all standard. This is the most expensive layer to add later.",[19,263,265],{"id":264},"takeaway","Takeaway",[11,267,268,269,272],{},"A reporting system requires far more decisions than it appears to. CityOS's module is a concrete application of the ",[142,270,271],{},"less clicking, more clarity"," principle.",{"title":71,"searchDepth":72,"depth":72,"links":274},[275,276,277,278,279],{"id":222,"depth":75,"text":223},{"id":229,"depth":75,"text":230},{"id":243,"depth":75,"text":244},{"id":257,"depth":75,"text":258},{"id":264,"depth":75,"text":265},"CityOS","2026-04-22","What questions did we ask while designing CityOS's reporting module? Three points where municipal workflows tend to break, and the answers we built.",{},"\u002Fen\u002Fblog\u002Fcityos-vatandas-bildirim-sistemi",7,{"title":211,"description":282},"en\u002Fblog\u002Fcityos-vatandas-bildirim-sistemi",[289,290,291],"public sector","product design","workflow","NPgZezqxZi8KgQ0XrRLvamU8IRsYrBEWgcwiJc-PNbg",1781523775268]