Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

70 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🧠 BRAHM-AI

License: BRAHM-AI Protected Platform Status API Status India First GitHub Stars GitHub Forks

BRAHM-AI is an India-first, Hindi-first intelligent ecosystem that connects Indian knowledge, multilingual conversation, live information, voice avatars, computer vision, automation, education, business intelligence, and practical digital tools through one unified platform.

It is designed to move beyond static chatbot responses:

Ask in your language. Understand with context. Act through connected intelligence.

BRAHM-AI combines ancient Indian knowledge with modern AI systems while maintaining a strong focus on responsible intelligence, data dignity, user control, and practical real-world utility.


🌐 Official Platform

brahm-ai.in is the primary entry point for the complete BRAHM-AI ecosystem. Voice, chat, language, Panchang, vision, scriptures, market intelligence, agriculture, education, automation, wallet services, and business AI products are progressively being unified under this platform.


✨ What Makes BRAHM-AI Different?

BRAHM-AI is not built as a single general-purpose chatbot. It uses a modular intelligence architecture in which different requests are routed to specialized systems.

A user may ask BRAHM-AI to:

  • Explain a verse from the Bhagavad Gita or Ramayan.
  • Provide today’s Panchang for a selected location.
  • Identify Tithi, Nakshatra, Rahu Kaal, Choghadiya, or Bhadra.
  • Understand an image, document, medicine strip, object, or live camera scene.
  • Talk through a multilingual voice avatar.
  • Translate or explain text in an Indian or international language.
  • Analyze recent news, market information, weather, or Mandi Bhav.
  • Automate supported communication and productivity tasks.
  • Receive education, Sanskrit, agriculture, and business assistance.
  • Use AI phone agents for customer calls, leads, and follow-ups.

The platform determines the appropriate capability and sends the request to the correct helper, computation engine, structured knowledge source, live-data provider, vision system, or reasoning layer.


🧩 Core Products

🎙️ BRAHM Voice

BRAHM Voice is the live voice-avatar experience of the ecosystem.

Key capabilities include:

  • Real-time Speech-to-Text interaction.
  • Natural Text-to-Speech responses.
  • Hindi-first and multilingual voice conversations.
  • Language-matched text, speech, and pronunciation.
  • Multiple avatar identities.
  • Listening, thinking, speaking, welcome, and idle states.
  • Lipsync and lightweight avatar animation.
  • Camera-assisted interaction where supported.
  • Returning-user and contextual welcome flows.
  • Voice-triggered supported actions.
  • Text fallback when speech output is unavailable.

BRAHM Voice does not maintain a separate knowledge engine. It uses the central BRAHM conversation and reasoning flow so that voice and text experiences remain consistent.


🧠 BRAHM True

BRAHM True is the primary deep-conversation and knowledge interface.

It coordinates:

  • General questions and contextual conversation.
  • Indian scriptures and philosophical knowledge.
  • Panchang and Jyotish requests.
  • Live news and web-supported information.
  • Market and business-related intelligence.
  • Vision and OCR observations.
  • User language and session context.
  • Specialized local helpers and structured data.
  • Safe model and service fallback paths.

BRAHM True acts as the central conversational source for connected BRAHM-AI experiences.


💬 BRAHM Samvad

BRAHM Samvad provides a language-focused conversational workspace.

It is designed for:

  • Multilingual conversations.
  • Translation assistance.
  • Grammar improvement.
  • Expression and sentence refinement.
  • Voice-assisted communication.
  • Partner-language interactions.
  • Text, voice, and supported scan-based translation.

🪔 Panchang and Jyotish Intelligence

BRAHM-AI includes dedicated Panchang and Jyotish-oriented capabilities.

Supported and developing capabilities include:

  • Tithi.
  • Nakshatra.
  • Yoga.
  • Karana.
  • Sunrise and sunset.
  • Moonrise and moonset.
  • Rahu Kaal.
  • Day and night Choghadiya.
  • Bhadra timing and contextual interpretation.
  • Hora.
  • Muhurta guidance.
  • Lahiri Ayanamsa.
  • Location-aware daily Panchang.
  • Birth-detail-based Kundli guidance.
  • Tara Bala and Chandra Bala workflows.
  • Natural-language Panchang questions.

Astronomical or calendar values should be produced through dedicated computation or structured data services before being explained by the conversational layer.

The language model is used for interpretation and explanation, not for inventing astronomical values.


📖 Scripture Intelligence

BRAHM-AI is being developed with a source-first Indian scripture architecture.

Knowledge areas include:

  • Bhagavad Gita.
  • Valmiki Ramayan.
  • Vedas.
  • Upanishads.
  • Puranas.
  • Sanskrit texts.
  • Dharma and Indian philosophy.
  • Mantras, Aartis, Stotras, and related literature.

The intended scripture pipeline is:

User Query → Scripture Detection → Exact Retrieval → Relevant Source Context → Explanation

The system prioritizes structured and locally indexed scripture data before generative interpretation.

Possible response elements include:

  • Scripture name.
  • Kanda, chapter, Sarga, or section.
  • Verse reference.
  • Original Sanskrit.
  • Hindi or English meaning.
  • Context.
  • Simple explanation.
  • Practical lesson.
  • Related source references.

🌍 Multilingual Intelligence

BRAHM-AI supports Hindi-first interaction while progressively expanding across Indian and international languages.

The multilingual pipeline aligns:

  • User-selected language.
  • Speech-to-Text language.
  • Script detection.
  • Input normalization.
  • Reasoning language.
  • Displayed response text.
  • Text-to-Speech language.
  • Pronunciation and supported accent.

Supported or developing language coverage includes major Indian languages and international languages such as English, Arabic, Persian, Hebrew, French, German, Spanish, Russian, Japanese, Korean, Thai, Indonesian, and others.

Language availability may vary between text, STT, TTS, translation, and voice-avatar modes.


📷 Vision Intelligence

BRAHM-AI Vision connects camera or uploaded visual input with contextual reasoning.

Supported and experimental flows include:

  • Image understanding.
  • Document reading.
  • OCR text extraction.
  • Object detection.
  • Multi-object recognition.
  • Scene understanding.
  • Animal recognition.
  • Traffic-light recognition.
  • Religious-place and environmental-scene recognition.
  • Medicine-strip and packaging reading.
  • Face-presence detection.
  • Returning-user recognition.
  • Mood and expression signals.
  • Approximate age-group estimation.
  • Camera-triggered follow-up questions.

The general vision pipeline is:

Camera or Image → Frame Validation → Detection or OCR → Structured Observation → BRAHM Reasoning → Text or Voice Response

Vision results may be affected by lighting, camera quality, angle, text clarity, model confidence, browser support, and device performance.


💊 Medicine Information Assistance

Users can show supported medicine packaging or strips to the camera and use commands such as:

  • “दवाई पढ़ो”
  • “दवाई देखो”
  • “Read this medicine”
  • “What medicine is this?”

The system may:

  1. Capture a stable frame.
  2. Extract visible text through OCR.
  3. Identify a likely medicine or composition.
  4. Present general informational details.
  5. Explain common uses, precautions, and known side effects.
  6. Ask the user to verify unclear packaging or consult a qualified professional.

Medicine identification is informational and may be incorrect when packaging is unclear or visually similar.

BRAHM-AI is not a medical device and does not replace a doctor, pharmacist, diagnosis, prescription, or emergency service.


📰 Live Intelligence

Where recent information is required, BRAHM-AI can route queries to live or recently updated sources.

Current and developing live-data areas include:

  • National and international news.
  • Technology and economic developments.
  • Weather.
  • Share-market context.
  • Company information.
  • Agricultural Mandi Bhav.
  • Location-aware utilities.
  • Time-sensitive public information.

Responses should distinguish between:

  • Structured local knowledge.
  • Live external data.
  • AI interpretation.
  • Information that could not be verified.

📈 Market Intelligence

BRAHM-AI includes market-focused analysis experiences for educational use.

Capabilities include:

  • Stock and crypto chart context.
  • Trend interpretation.
  • Support and resistance observations.
  • Momentum and volatility signals.
  • News-supported market context.
  • Bullish and bearish scenarios.
  • Risk-focused explanations.
  • Company and sector understanding.

Market analysis is educational and is not financial or investment advice.


🌾 Agriculture and Mandi Intelligence

BRAHM-AI provides agriculture-oriented assistance through natural-language interaction.

Supported and developing areas include:

  • City and crop-specific Mandi Bhav.
  • Weather-aware farming context.
  • Seasonal crop guidance.
  • Rainfall and humidity context.
  • Crop-risk signals.
  • Government and public-data integrations.
  • Hindi-first accessibility.
  • Voice-based information delivery.

Live agricultural information depends on the availability, freshness, reliability, and response time of external data providers.


🎓 Education and Language Tools

The BRAHM-AI ecosystem includes specialized education modules such as:

  • Aryabhata: Mathematics learning and step-by-step explanations.
  • Panini: Sanskrit grammar, Sutra discovery, and language learning.
  • Vedavyas Library: Indian scripture and knowledge resources.
  • BRAHM Teach: School-level educational assistance.
  • Translation and grammar tools.
  • Voice-supported lessons.
  • Structured study and practice workflows.

🎮 Indian Games and Cultural Experiences

BRAHM-AI also preserves and presents Indian strategic and spiritual games.

Additional interactive experiences include VR environments, cultural learning, spiritual media, and Indian knowledge-based digital exploration.


🪙 PotliPay and Ecosystem Access

PotliPay is the account, balance, and AI-usage layer of the BRAHM-AI ecosystem.

Its architecture is designed to support:

  • BRAHM account identity.
  • AI usage balances.
  • Free and purchased usage allowances.
  • BHM-based ecosystem utility.
  • Transaction and usage history.
  • Product access.
  • Future API usage metering.
  • Plan-based capability access.

Blockchain or token-related functionality may be introduced progressively and is subject to product, technical, compliance, and network requirements.


📞 BRAHM AI Agents

BRAHM AI Agents is the business-focused AI phone-agent platform.

It is designed for:

  • Inbound customer calls.
  • Outbound workflows.
  • Lead capture.
  • Follow-up conversations.
  • Multilingual call handling.
  • CRM timelines.
  • Call summaries.
  • Appointment and enquiry workflows.
  • Business knowledge integration.
  • Usage and billing controls.

Business platform:

www.brahm-ai-agents.com

Application dashboard:

app.brahm-ai-agents.com


🏗️ Technical Architecture

BRAHM-AI uses a modular, API-oriented architecture.

High-Level Flow

User
  │
  ├── Text
  ├── Voice
  ├── Camera / Image
  ├── Document
  └── Supported Action Request
        │
        ▼
Input Normalization Layer
        │
        ├── Language Detection
        ├── Speech-to-Text
        ├── OCR
        ├── Vision Observations
        └── Session and User Context
        │
        ▼
Intent and Capability Router
        │
        ├── General Conversation
        ├── Scripture Retrieval
        ├── Panchang Computation
        ├── Vision Intelligence
        ├── Live Search and News
        ├── Market Intelligence
        ├── Mandi and Agriculture
        ├── Translation and Language
        └── Automation
        │
        ▼
BRAHM Core Reasoning Layer
        │
        ├── Exact-Source Response
        ├── Retrieval-Augmented Response
        ├── Live-Data Explanation
        ├── General Reasoning
        └── Safe Fallback
        │
        ▼
Response Delivery Layer
        │
        ├── Text
        ├── Multilingual TTS
        ├── Voice Avatar
        ├── Structured Data
        └── Supported Action Result

🧱 Architecture Principles

1. Specialized Routing

A single model is not expected to perform every task.

BRAHM-AI routes requests to the most appropriate:

  • Helper.
  • Database.
  • Search provider.
  • Calculation service.
  • Vision model.
  • OCR engine.
  • Scripture index.
  • Market-data provider.
  • Automation connector.
  • Reasoning model.

2. Exact Data Before Generation

For structured domains such as Panchang, scriptures, prices, user balances, or account information:

  1. Retrieve or compute the value.
  2. Validate the available result.
  3. Pass structured context to the reasoning layer.
  4. Generate a clear user-facing explanation.

The model should not invent exact values when a dedicated data source is required.


3. Shared Intelligence Flow

BRAHM Voice and BRAHM True use a connected intelligence flow.

The Voice interface handles:

  • Microphone input.
  • STT.
  • Avatar states.
  • TTS.
  • Lipsync.
  • Voice interaction.

The central chat and reasoning flow remains responsible for selecting helpers, retrieving knowledge, and producing the actual response.


4. API-Driven Frontend

The long-term platform architecture keeps heavy logic away from the user-interface layer.

Frontend responsibilities include:

  • User interaction.
  • Input capture.
  • State presentation.
  • Lightweight animation.
  • Accessibility.
  • Response display.

Backend/API responsibilities include:

  • Authentication.
  • Capability routing.
  • Knowledge retrieval.
  • Computation.
  • Model access.
  • Usage metering.
  • Memory.
  • Logging.
  • Security enforcement.
  • Response formatting.

5. Modular Deployment

Major modules remain independently maintainable.

Examples:

  • Voice.
  • BRAHM True.
  • Vision.
  • Language.
  • Panchang.
  • Jyotish.
  • Scripture.
  • News.
  • Market.
  • Agriculture.
  • Automation.
  • Wallet.
  • Business Agents.

This allows one capability to be upgraded without rebuilding the complete ecosystem.


6. Fallback-Oriented Reliability

A supported request may use multiple response levels:

  1. Exact local helper.
  2. Structured database.
  3. Retrieval-Augmented Generation.
  4. Live provider.
  5. Primary reasoning model.
  6. Secondary model or service.
  7. Source-only response.
  8. Safe text fallback.

The system should provide an honest failure response instead of fabricating unavailable information.


🔌 BRAHM-AI Developer API — Coming Soon

A public and partner API layer is under development.

The API is intended to allow websites, applications, institutions, developers, and businesses to access selected BRAHM-AI capabilities without embedding the complete user interface.

Planned API Categories

  • General BRAHM conversation.
  • Multilingual language processing.
  • Panchang and Muhurta.
  • Jyotish utilities.
  • Scripture search and retrieval.
  • Sanskrit assistance.
  • Vision and OCR.
  • Medicine-label reading.
  • Market intelligence.
  • Mandi Bhav and agriculture.
  • Translation.
  • Voice sessions.
  • Text-to-Speech.
  • Automation.
  • Business-agent integrations.

🛣️ Draft API Surface

The following endpoints represent the planned API direction and may change before public release.

Core Conversation

POST /api/v1/chat

Example request:

{
  "message": "आज जयपुर का पंचांग बताइए",
  "language": "hi-IN",
  "location": {
    "city": "Jaipur",
    "country": "India"
  },
  "session_id": "optional-session-id"
}

Panchang

GET /api/v1/panchang/daily

Possible parameters:

date=2026-07-16
latitude=26.9124
longitude=75.7873
timezone=Asia/Kolkata
language=hi

Potential response fields:

{
  "date": "2026-07-16",
  "location": "Jaipur, India",
  "timezone": "Asia/Kolkata",
  "tithi": {},
  "nakshatra": {},
  "yoga": {},
  "karana": {},
  "sunrise": "",
  "sunset": "",
  "rahu_kaal": {},
  "choghadiya": [],
  "bhadra": {},
  "calculation_source": "",
  "generated_at": ""
}

Scripture Search

POST /api/v1/scriptures/search

Example request:

{
  "query": "विराध कौन था?",
  "scripture": "valmiki_ramayan",
  "language": "hi",
  "include_original_text": true,
  "max_results": 5
}

Potential response:

{
  "answer": "",
  "references": [
    {
      "scripture": "Valmiki Ramayan",
      "section": "Aranyakanda",
      "chapter": 2,
      "verse": 12,
      "sanskrit": "",
      "meaning": ""
    }
  ],
  "retrieval_mode": "exact"
}

Vision and OCR

POST /api/v1/vision/analyze

Planned analysis modes:

object
scene
document
ocr
medicine
general

Example multipart request:

image: <uploaded file>
mode: medicine
language: en

Potential response:

{
  "mode": "medicine",
  "detected_text": [],
  "possible_name": "",
  "confidence": 0.0,
  "explanation": "",
  "warnings": [],
  "requires_user_verification": true
}

Translation and Language

POST /api/v1/language/process

Example request:

{
  "text": "आप कैसे हैं?",
  "source_language": "auto",
  "target_language": "fa",
  "task": "translate"
}

Potential tasks:

detect
translate
normalize
explain
grammar_check
transliterate

Mandi Bhav

GET /api/v1/mandi/prices

Example parameters:

crop=wheat
city=Jaipur
state=Rajasthan
language=hi

Market Intelligence

POST /api/v1/market/analyze

Example request:

{
  "symbol": "NSE:RELIANCE",
  "timeframe": "1D",
  "currency": "INR",
  "language": "en"
}

Voice Session

POST /api/v1/voice/session

The voice API is expected to coordinate:

  • Session initialization.
  • Language selection.
  • STT configuration.
  • Conversation routing.
  • TTS voice selection.
  • Audio delivery.
  • Usage metering.

🔐 Planned API Authentication

The public API is expected to use API-key or bearer-token authentication.

Example:

Authorization: Bearer brahm_live_xxxxxxxxx
Content-Type: application/json

Planned key environments:

  • Development or test keys.
  • Production keys.
  • Restricted partner keys.
  • Internal service keys.

API keys must never be exposed directly inside public frontend code.


📦 Standard API Response Envelope

Planned success format:

{
  "success": true,
  "request_id": "req_xxxxx",
  "data": {},
  "usage": {
    "units": 1,
    "remaining": 999
  },
  "meta": {
    "language": "hi",
    "source_mode": "structured",
    "generated_at": "2026-07-16T12:00:00Z"
  }
}

Planned error format:

{
  "success": false,
  "request_id": "req_xxxxx",
  "error": {
    "code": "INVALID_REQUEST",
    "message": "A valid location is required."
  }
}

⚙️ Planned API Platform Features

  • Versioned endpoints.
  • API-key management.
  • Authentication and authorization.
  • Per-key quotas.
  • Rate limiting.
  • Usage metering.
  • Request IDs.
  • Structured error codes.
  • Dashboard analytics.
  • Development and production environments.
  • Configurable language output.
  • Source and freshness metadata.
  • Streaming responses where supported.
  • Webhooks for supported automation events.
  • Partner-level permissions.
  • Usage-based BHM or plan billing.
  • Abuse and duplicate-request protection.
  • Audit and operational logs.

🛡️ Responsible AI and Safety

BRAHM-AI is designed around responsible use.

Knowledge Integrity

  • Structured sources should be preferred for exact facts.
  • Generated explanations should not be presented as original scripture.
  • Live information should include source freshness where possible.
  • Unverified information should be identified clearly.

User Control

  • Sensitive actions should require user confirmation.
  • Camera and microphone access require browser or device permission.
  • Users should understand when live data or external services are being used.
  • Personalization should remain limited to relevant user context.

Medical Safety

  • Medicine and health-related outputs are informational only.
  • Camera or OCR recognition can be wrong.
  • BRAHM-AI does not provide medical diagnosis or prescriptions.
  • Urgent health concerns require qualified medical or emergency support.

Financial Safety

  • Market analysis is educational.
  • Price data may be delayed or incomplete.
  • BRAHM-AI does not guarantee returns or financial outcomes.

Spiritual and Cultural Responsibility

  • Scripture references should be handled respectfully.
  • Textual variants and editions may differ.
  • Explanations should distinguish source text from interpretation.
  • The platform is an educational and guidance system, not a replacement for a qualified traditional teacher or Acharya.

📱 Platform Experience

BRAHM-AI is designed for:

  • Mobile and desktop browsers.
  • Hindi-first interaction.
  • Voice, text, camera, and touch input.
  • Progressive Web App installation.
  • Lightweight avatar animation.
  • Modular page loading.
  • Session continuity.
  • Accessible text fallback.
  • Account-linked personalization.

Browser and device support may differ for microphone, camera, speech recognition, speech synthesis, and background media behaviour.


📂 Repository Purpose

This repository currently serves as a public development and documentation repository.

It contains or may contain:

  • Project overview.
  • Architecture summaries.
  • Public technical documentation.
  • Daily development updates.
  • Feature histories.
  • API plans.
  • Integration examples.
  • Product roadmaps.
  • Selected public resources.

The complete production source code, private models, datasets, credentials, internal routing logic, infrastructure configuration, and protected commercial modules are not necessarily included.

Daily updates are available in:

UPDATES.md


🤝 Collaboration and Licensing

BRAHM-AI welcomes discussions with:

  • AI developers.
  • Researchers.
  • Sanskrit and scripture scholars.
  • Educational institutions.
  • Agriculture and public-data organizations.
  • Healthcare-information specialists.
  • Indian-language contributors.
  • Businesses seeking AI integrations.
  • API and infrastructure partners.

Collaboration, source access, dataset access, API access, model access, deployment rights, and commercial usage may require a separate written agreement or license.


🗺️ Development Roadmap

Current Platform

  • Unified brahm-ai.in ecosystem.
  • BRAHM True.
  • BRAHM Voice.
  • Multilingual interaction.
  • Vision and OCR.
  • Panchang and Jyotish utilities.
  • Scripture intelligence.
  • Live information modules.
  • Market and agriculture intelligence.
  • PotliPay ecosystem access.
  • BRAHM AI Agents.

In Development

  • Public BRAHM-AI API.
  • Developer dashboard.
  • API-key and quota management.
  • Expanded structured scripture datasets.
  • Improved multilingual STT and TTS.
  • Advanced local intelligence models.
  • Better source attribution.
  • Expanded automation connectors.
  • Business and institutional API plans.
  • Lightweight on-device and local inference options.
  • Improved memory and personalized context controls.

Future Direction

  • Partner SDKs.
  • Streaming voice API.
  • Embeddable website assistant.
  • Panchang widgets.
  • Scripture-search API.
  • Vision and document API.
  • Business-agent integrations.
  • Advanced Indian-language intelligence.
  • Institution-specific knowledge systems.
  • Secure private deployment.
  • Local and sovereign AI infrastructure.

🌐 Quick Links

Experience Link
Main Platform brahm-ai.in
BRAHM Voice brahm-ai.in/voice
BRAHM True brahm-ai.in/true
Chaturang brahm-ai.in/chaturang.php
Adu-Puli brahm-ai.in/adu-puli.php
Moksha Path brahm-ai.in/moksha-path.php
Games Hub brahm-ai.in/games-hub.php
Cosmic Clock brahm-ai.in/cosmic-clock.php
BRAHM AI Agents brahm-ai-agents.com
Development Updates UPDATES.md

🙏 Credits

BRAHM-AI is:

  • Rooted in Indian knowledge traditions.
  • Built with a Hindi-first and multilingual vision.
  • Developed as part of the Royal Web 3 Blockchain and BRAHM-AI ecosystem.
  • Supported by structured knowledge, modern AI, live-data services, voice systems, vision systems, and modular software architecture.
  • Developed with the objective of connecting knowledge, intelligence, and useful action.

⭐ Community Support

You can support the project by:

  • Starring this repository.
  • Following development updates.
  • Reporting documentation issues.
  • Suggesting responsible use cases.
  • Contributing language or domain knowledge through approved collaboration.
  • Sharing BRAHM-AI with developers, researchers, institutions, and users.

📄 License

This repository and the BRAHM-AI name, documentation, designs, architecture, datasets, models, workflows, and associated intellectual property are protected.

Unless explicitly stated otherwise:

  • No production or commercial usage rights are granted.
  • No redistribution or reselling rights are granted.
  • No model, dataset, branding, or source-code rights are granted.
  • Forking the public repository does not grant a commercial license.
  • Written permission may be required for integration, deployment, modification, or derivative commercial use.

For licensing, partnership, or authorized access, contact the BRAHM-AI team through the official website.


BRAHM-AI — Ancient Knowledge. Modern Intelligence. Connected Action.

© 2025–2026 BRAHM-AI and Royal Web 3 Blockchain. All Rights Reserved.

About

BRAHM-AI is an India-first multilingual AI ecosystem for voice avatars, vision, Sanskrit, Panchang, scripture RAG, automation, local LLMs and developer APIs.

Topics

Resources

Contributing

Security policy

Stars

2 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors