# Shodh > Shodh is a private, on-premise enterprise AI platform for Indian industry, built by Corp8 AI (Corpeight Labs Pvt. Ltd., Ahmedabad, Gujarat). It organizes documents, answers employee questions from company sources with citations, manages SOPs and training, extracts data from scanned records, and automates workflows — with document processing, search indexing and AI inference running inside the customer's own infrastructure. No customer content is sent to external AI services during operation. Status: private beta (2026). ## Modules - Document Intelligence: ingestion (PDF, Word, Excel, PPT, email, scans), OCR, classification, tagging, metadata, duplicate detection, versioning - Knowledge Assistant: permission-aware hybrid search, cited answers, file comparison, summaries, expert knowledge capture - SOP & Training: SOP drafting, templates, version comparison, approval workflows, effective dates, checklists, onboarding guides, quizzes - Operational Intelligence: maintenance assistance from manuals and service history, shift handovers, incident lookup, record review - Data & Model Adaptation: data cleaning and standardization, reports from approved databases, local model management, fine-tuning when justified - Workflow Automation: triggers, approval routing, notifications, ERP/system updates with human review gates - Privacy & Administration: SSO, RBAC, document-level access, encryption, audit logs, sensitive-data detection and redaction, retention/backup/deletion ## Industries Manufacturing & engineering; pharmaceuticals & laboratories; food processing & packaging; logistics & warehousing; government & public bodies; infrastructure & construction; professional services. ## India focus English, Hindi, Gujarati and other Indian languages (validated per deployment); mixed-language (Hinglish) queries; scanned paper archives; data sovereignty on customer premises; air-gapped deployment option. ## Deployment Discovery (week 1) → configuration (weeks 2–4) → validation on real tasks (weeks 4–6) → rollout. Tiers: Pilot (one department), Enterprise (organisation-wide, SSO/ERP integrations, HA), Sovereign (air-gapped / government). Hardware sized to workload, from a single GPU server upward. Pricing scoped per deployment. ## Key facts - Company knowledge does not require fine-tuning; Shodh uses retrieval-augmented generation and re-indexes updated documents. - Employees only retrieve information their permissions allow. - SOPs are published only after designated owners approve them. ## Links - Website: https://xn--shdh-wqa.com/ (shôdh.com) - Product app: https://shodh.corp8.ai - API: https://shodhapi.corp8.ai - Beta access: https://xn--shdh-wqa.com/#beta - FAQs: https://xn--shdh-wqa.com/#faq - Company: https://corp8.ai - Contact: niraj.ojha@corp8.ai ## Blog - [Air-Gapped AI Deployment in India: A Practical Guide for Pharma, Manufacturing and Other Regulated Workloads](https://xn--shdh-wqa.com/blog/air-gapped-ai-deployment-india-guide/): Air-gapped AI for Indian pharma, manufacturing and logistics: isolation tiers, DPDP and GxP drivers, stack components, costs and a 90-day pilot plan. - [RAG vs. AI Agents: What Indian Manufacturers Should Build First for On-Premise Knowledge Automation](https://xn--shdh-wqa.com/blog/rag-vs-ai-agents-on-premise-knowledge-automation/): RAG retrieves answers from your own documents; AI agents take actions. Here's which one Indian manufacturers should build first on-premise. - [Cloud Repatriation for AI in India: Which Workloads to Move On-Premise and How to Decide](https://xn--shdh-wqa.com/blog/cloud-repatriation-for-ai-india/): A practical guide for Indian manufacturers, pharma, logistics and public-sector teams on which AI workloads to move on-premise — and how to decide. - [83% of Enterprises Are Repatriating AI Workloads: What It Means for On-Premise AI in India](https://xn--shdh-wqa.com/blog/on-premise-ai-india-repatriation/): Broadcom found 83% of enterprises plan to repatriate AI workloads. What that shift means for on-premise AI in India: costs, DPDP, and a pilot playbook. - [Google Brings Air-Gapped Gemini to India: What Regulated Enterprises Should Weigh Before Choosing Private AI](https://xn--shdh-wqa.com/blog/on-premise-ai-india-air-gapped-gemini/): Google's air-gapped Gemini makes private AI mainstream in India. Here is how air-gapped cloud differs from on-premise AI on custody, compliance and cost. - [From Data Localisation to AI Localisation: How Indian Enterprises Can Run AI Inference On-Premise for DPDP Compliance](https://xn--shdh-wqa.com/blog/on-premise-enterprise-ai-dpdp-compliance-india/): Why Indian enterprises must localise AI processing, not just data storage, for DPDP compliance — and how on-premise AI inference delivers it. - [Why Indian Executives Fear Data Sovereignty Failures — and How On-Premise AI Fixes Them](https://xn--shdh-wqa.com/blog/on-premise-ai-india-data-sovereignty/): Nine in ten Indian executives fear a data sovereignty failure could cost their job. On-premise AI keeps documents, models and audit trails in your control. - [Why 90% of Indian Executives Fear Data Sovereignty Failures — and the On-Premise AI Checklist That Fixes It](https://xn--shdh-wqa.com/blog/on-premise-ai-india-data-sovereignty-checklist/): 90% of Indian executives link data sovereignty failures to job losses. Here is the on-premise AI checklist that turns that fear into verifiable control. - [On-Premise AI for Indian Manufacturers: What It Is, What It Costs and Where to Start](https://xn--shdh-wqa.com/blog/on-premise-ai-for-indian-manufacturers/): A practical guide to on-premise AI for Indian manufacturers: what it costs, how DPDP and data sovereignty shape the choice, and a 90-day pilot plan.