{
  "name": "Piyush Bodaa",
  "url": "https://www.piyushbodaa.com/",
  "location": {
    "locality": "Hyderabad",
    "country": "IN",
    "label": "Hyderabad, India"
  },
  "description": "Tools for learning, privacy, and public accountability, built by Piyush Bodaa in Hyderabad.",
  "headline": "I build tools for learning, privacy, and public accountability.",
  "summary": "Right now that’s Krutrim Mulyankan, a free app that shows which AI answers your own questions best, and what each answer costs in rupees. I also run Data Libertarian: public registers of the Indian state, every entry linked to its source.",
  "email": "email@piyushbodaa.com",
  "sameAs": [
    "https://x.com/piyushbodaa"
  ],
  "pages": {
    "home": "https://www.piyushbodaa.com/",
    "work": "https://www.piyushbodaa.com/work",
    "llms": "https://www.piyushbodaa.com/llms.txt",
    "fullText": "https://www.piyushbodaa.com/llms-full.txt"
  },
  "openSource": {
    "title": "Open source, *where I’m contributing next.*",
    "intro": "I’m starting to contribute to open-source software, mostly to two projects.",
    "github": "https://github.com/piyushbodaa",
    "projects": [
      {
        "name": "Omarchy",
        "kind": "Linux desktop · Shell · MIT",
        "text": "Beautiful, modern and opinionated Linux, built on Arch and Hyprland. It’s the desktop I build on every day.",
        "url": "https://omarchy.org",
        "repo": "https://github.com/omacom/omarchy"
      },
      {
        "name": "Hermes Agent",
        "kind": "AI agent · Python · MIT",
        "text": "Nous Research’s open-source AI agent: “the agent that grows with you”.",
        "url": "https://hermes-agent.nousresearch.com",
        "repo": "https://github.com/NousResearch/hermes-agent"
      }
    ]
  },
  "lists": {
    "projects": "Complete portfolio, in the order shown on /work.",
    "featured": "Project ids shown under Selected work on the homepage, in order.",
    "work": "Data Libertarian and Krutrim Mulyankan plus every project in the apps and companies categories.",
    "categories": "Project names grouped by kind."
  },
  "featured": [
    "krutrim-mulyankan",
    "data",
    "gta-nanded",
    "hyderabad-reads",
    "otgs"
  ],
  "work": [
    {
      "id": "krutrim-mulyankan",
      "category": "websites",
      "name": "Krutrim Mulyankan",
      "nativeName": "कृत्रिम मूल्यांकन",
      "kicker": "AI evaluation",
      "tagline": "AI evaluations for India",
      "summary": "A free desktop app for finding the AI that works for you. Export your chat history, let the app find the answers your AI got wrong, and turn them into a personal test that every other model can take — with the cost of each answer in rupees.",
      "detail": "The site also explains six published Indian-language AI benchmarks, with their dates and limits, and proposes 300 everyday tasks for evaluations of its own.",
      "facts": [
        {
          "value": "300",
          "label": "Everyday tasks proposed"
        },
        {
          "value": "6",
          "label": "Indian AI benchmarks explained"
        }
      ],
      "url": "https://krutrimmulyankan.in",
      "status": "In development",
      "state": "building",
      "platform": "Desktop app for Linux; website in English and Hindi",
      "imageAlt": "Krutrim Mulyankan’s homepage: “Find the AI that works for you”, above a short film about the desktop app.",
      "tone": "dark"
    },
    {
      "id": "data",
      "category": "websites",
      "name": "Data Libertarian",
      "kicker": "Public records",
      "tagline": "Public registers of the Indian state",
      "summary": "Public registers of the Indian state: court findings against public servants, prosecutions for speech and for crimes with no victim, penalties on peaceful economic activity, the businesses the state owns and the money it spends. Every entry names the court or agency that decided it and links to its source.",
      "detail": "It also reads official books line by line — budgets from the Union down to villages, state police budgets, price indices and the National Crime Records Bureau’s tables for 2022 to 2024. An empty cell means a figure has not been read yet, never zero.",
      "facts": [
        {
          "value": "2,697",
          "label": "Case records on four registers"
        },
        {
          "value": "135",
          "label": "Public enterprises inventoried"
        }
      ],
      "factsSource": "Counted on datalibertarian.in on 21 October 2026.",
      "url": "https://datalibertarian.in",
      "status": "Live",
      "state": "live",
      "platform": "Website",
      "imageAlt": "Data Libertarian’s homepage: “The Indian state, on the record”, beside a panel counting the case records on each register.",
      "tone": "light"
    },
    {
      "id": "libby",
      "category": "apps",
      "name": "Libby",
      "kicker": "On-device AI",
      "tagline": "A private AI that lives on your phone",
      "summary": "A private AI that runs entirely on your phone. Download a small open model once, then chat offline — no account, no cloud, and nothing leaves the device.",
      "detail": "It runs models on the phone with llama.cpp, lets you swap GGUF models whenever you like, and comes dressed in gold and black, with a small snake for company.",
      "status": "In development",
      "state": "building",
      "platform": "Android",
      "notify": "Not released yet. Email me to hear when it is.",
      "imageAlt": "Libby’s welcome screen: a gold snake above three promises — on-device, private, and yours.",
      "tone": "dark"
    }
  ],
  "projects": [
    {
      "id": "krutrim-mulyankan",
      "category": "websites",
      "name": "Krutrim Mulyankan",
      "nativeName": "कृत्रिम मूल्यांकन",
      "kicker": "AI evaluation",
      "tagline": "AI evaluations for India",
      "summary": "A free desktop app for finding the AI that works for you. Export your chat history, let the app find the answers your AI got wrong, and turn them into a personal test that every other model can take — with the cost of each answer in rupees.",
      "detail": "The site also explains six published Indian-language AI benchmarks, with their dates and limits, and proposes 300 everyday tasks for evaluations of its own.",
      "facts": [
        {
          "value": "300",
          "label": "Everyday tasks proposed"
        },
        {
          "value": "6",
          "label": "Indian AI benchmarks explained"
        }
      ],
      "url": "https://krutrimmulyankan.in",
      "status": "In development",
      "state": "building",
      "platform": "Desktop app for Linux; website in English and Hindi",
      "imageAlt": "Krutrim Mulyankan’s homepage: “Find the AI that works for you”, above a short film about the desktop app.",
      "tone": "dark"
    },
    {
      "id": "data",
      "category": "websites",
      "name": "Data Libertarian",
      "kicker": "Public records",
      "tagline": "Public registers of the Indian state",
      "summary": "Public registers of the Indian state: court findings against public servants, prosecutions for speech and for crimes with no victim, penalties on peaceful economic activity, the businesses the state owns and the money it spends. Every entry names the court or agency that decided it and links to its source.",
      "detail": "It also reads official books line by line — budgets from the Union down to villages, state police budgets, price indices and the National Crime Records Bureau’s tables for 2022 to 2024. An empty cell means a figure has not been read yet, never zero.",
      "facts": [
        {
          "value": "2,697",
          "label": "Case records on four registers"
        },
        {
          "value": "135",
          "label": "Public enterprises inventoried"
        }
      ],
      "factsSource": "Counted on datalibertarian.in on 21 October 2026.",
      "url": "https://datalibertarian.in",
      "status": "Live",
      "state": "live",
      "platform": "Website",
      "imageAlt": "Data Libertarian’s homepage: “The Indian state, on the record”, beside a panel counting the case records on each register.",
      "tone": "light"
    },
    {
      "id": "gta-nanded",
      "category": "websites",
      "name": "GTA Nanded",
      "kicker": "Game",
      "tagline": "An open-world game of Nanded",
      "summary": "An open-world browser game set in Nanded, Maharashtra. Drive autos, Lal Pari buses and police jeeps through the city’s real street map, run fares, and outrun Inspector Kadam on the way from Hazur Sahib station to Nandgiri Fort.",
      "detail": "Eight story missions or free roam, with progress saved on your device. It plays with a keyboard, a gamepad or touch, on map data from OpenStreetMap.",
      "facts": [
        {
          "value": "8",
          "label": "Story missions"
        }
      ],
      "url": "https://gta.nanded.fun",
      "status": "Live",
      "state": "live",
      "platform": "Browser game, desktop and phone",
      "imageAlt": "GTA Nanded’s title screen: an illustrated poster of the city and its riders beside the menu for a new story, free roam and the intro.",
      "tone": "dark"
    },
    {
      "id": "hyderabad-reads",
      "category": "websites",
      "name": "Hyderabad Reads",
      "kicker": "Book club",
      "tagline": "A book club’s shared library",
      "summary": "The shared bookshelf and meetup calendar of a weekend book club in Hyderabad. Members list the books they’re happy to lend or swap; find one you like and ask its owner to bring it to the next meetup.",
      "detail": "It is open source under the MIT licence. Members add their books on GitHub or send me the titles, and covers and dates come from Open Library.",
      "facts": [
        {
          "value": "95",
          "label": "Books on the shelves"
        },
        {
          "value": "19",
          "label": "Readers"
        }
      ],
      "factsSource": "Counted on hyderabadreads.fun on 21 October 2026.",
      "url": "https://hyderabadreads.fun",
      "status": "Live",
      "state": "live",
      "platform": "Website",
      "imageAlt": "Hyderabad Reads’ homepage: “Every book on our shelves, in one place.”, beside a stack of book covers and counts of books, readers and genres.",
      "tone": "dark"
    },
    {
      "id": "otgs",
      "category": "websites",
      "name": "Oxford The Global School",
      "kicker": "School website",
      "tagline": "A CBSE school in Nanded",
      "summary": "The website of Oxford The Global School, a co-educational CBSE school in Nanded, Maharashtra, from Pre-Primary to Grade 10: admissions, academics, life at the school and the CBSE mandatory disclosures in one place.",
      "detail": "It covers admissions for 2026–27, the curriculum, clubs and houses, the summer camp and a gallery of the campus.",
      "url": "https://otgs.fun",
      "status": "Live",
      "state": "live",
      "platform": "Website",
      "imageAlt": "Oxford The Global School’s homepage: “Rooted in values. Ready for the world.” over a photo of the school building, with admissions open for 2026–27.",
      "tone": "dark"
    },
    {
      "id": "typeflow",
      "category": "websites",
      "name": "Typeflow",
      "kicker": "Typing trainer",
      "tagline": "A typing trainer set to poetry",
      "summary": "A typing trainer where you type poetry, not random words. Every test and lesson is a stanza from a well-loved poem — Shakespeare to Tagore, Basho to Kabir — credited on screen.",
      "detail": "Timed tests, twelve lessons in sequence, and a record of your speed and accuracy that stays in your browser. The clock waits until you begin.",
      "url": "https://typeflow.piyushbodaa.com",
      "status": "Live",
      "state": "live",
      "platform": "Website, best with a physical keyboard",
      "imageAlt": "Typeflow’s homepage: “Sit down. Type the next line.”, with buttons for a timed test, lessons and history.",
      "tone": "dark"
    },
    {
      "id": "wordflow",
      "category": "websites",
      "name": "Wordflow",
      "kicker": "Vocabulary",
      "tagline": "Vocabulary tests, then practice",
      "summary": "Fifteen-question vocabulary tests, with no clock, in everyday English or the language of AI models — then five minutes of practice. No account required.",
      "url": "https://english.piyushbodaa.com",
      "status": "Live",
      "state": "live",
      "platform": "Website",
      "imageAlt": "Wordflow’s homepage: “Take the test. Then improve.”, with a choice of an English test or an AI test.",
      "tone": "light"
    },
    {
      "id": "dropwell",
      "category": "websites",
      "name": "Dropwell",
      "kicker": "Game",
      "tagline": "A falling-block puzzle",
      "summary": "A falling-block puzzle: drop glyphs, clear lines, and don’t fill the well.",
      "detail": "I built it with Grok Build after an evening of Tetris. It is also how this site got started.",
      "url": "https://dropwell.piyushbodaa.com",
      "status": "Live",
      "state": "live",
      "platform": "Website",
      "imageAlt": "Dropwell’s start screen: an empty well with a Play button, between panels for the held piece, score, lines, level and the next pieces.",
      "tone": "dark"
    },
    {
      "id": "libby",
      "category": "apps",
      "name": "Libby",
      "kicker": "On-device AI",
      "tagline": "A private AI that lives on your phone",
      "summary": "A private AI that runs entirely on your phone. Download a small open model once, then chat offline — no account, no cloud, and nothing leaves the device.",
      "detail": "It runs models on the phone with llama.cpp, lets you swap GGUF models whenever you like, and comes dressed in gold and black, with a small snake for company.",
      "status": "In development",
      "state": "building",
      "platform": "Android",
      "notify": "Not released yet. Email me to hear when it is.",
      "imageAlt": "Libby’s welcome screen: a gold snake above three promises — on-device, private, and yours.",
      "tone": "dark"
    }
  ],
  "categories": {
    "websites": [
      "Krutrim Mulyankan",
      "Data Libertarian",
      "GTA Nanded",
      "Hyderabad Reads",
      "Oxford The Global School",
      "Typeflow",
      "Wordflow",
      "Dropwell"
    ],
    "apps": [
      "Libby"
    ],
    "companies": {
      "nonProfit": []
    }
  }
}