Key Tender Details

Important Dates

Scope of Work

The selected technology partner was required to design, develop, deploy and maintain an AI-powered semantic search and discovery platform for the Ministry of Culture’s digital ecosystem. The solution was expected to provide intelligent content retrieval from culture.gov.in and approximately 30 other Ministry-related websites and platforms.

Key scope components include:

1. Development of an advanced semantic vector search engine integrated with the Ministry of Culture website.
2. Crawling, indexing and retrieval of content including text, images, audio, video and metadata.
3. LLM-powered search that goes beyond keyword matching and understands user intent, context and relevance.
4. AI-generated summaries and contextual responses using Retrieval-Augmented Generation architecture.
5. AI-driven chatbot for guiding users and enabling natural language interaction.
6. Integration with structured and unstructured data sources such as PDFs, Word documents, Excel files, APIs, cloud drives and web pages.
7. Bilingual support in English and Hindi for search queries, translation, transliteration and summarisation.
8. Regular model training, fine-tuning and monthly model retraining cycles.
9. Analytics panel for user query analysis and service improvement.
10. Security audit, encryption, compliance checks, auditability and traceability.
11. Deployment on scalable infrastructure, preferably in a MeitY-empanelled Indian data centre.
12. Ongoing support, troubleshooting, maintenance and monthly model updates for the contract duration.

Infrastructure and Technical Requirements

The bidder was required to provide GPU-based high-performance computing infrastructure for AI model training and inference. The minimum infrastructure expectations included multi-core CPU servers, minimum 128 GB RAM, high-speed SSD storage with at least 10 TB capacity, GPU-enabled servers, redundant network setup, backup and disaster recovery, high availability, and capacity to support projected concurrent users across around 30 ring-fenced websites.

The solution was also required to maintain 99.5% uptime, respond to 95% of queries within less than 3 seconds, and achieve at least 95% accuracy based on sample verification and RAG source validation.

Pre-Qualification and Eligibility Criteria

Eligible bidders were required to be registered legal entities in India, including Government organisations, PSUs, private limited companies, partnership firms, corporate entities under Indian laws, or autonomous institutions of the Government of India. Consortiums and joint ventures were not allowed.

Bidders were also required to submit an undertaking confirming that they had not been blacklisted or debarred by public financial institutions, public sector banks, RBI, IBA, corporate entities or any government agency during the last five years.

The bidder had to demonstrate relevant experience in similar IT/ITES works such as website or portal development, AI/ML integration, LLM-based search, semantic search, chatbot systems, vector databases, content summarisation and AI/ML/LLM integration.

Financial Criteria

The bidder was required to have an average annual turnover of at least Rs. 25 crore from IT/ITES activities over the last three financial years: 2022-23, 2023-24 and 2024-25. A certificate from a Chartered Accountant, along with audited financial statements, was required as supporting documentation.

For technical evaluation, turnover was also scored as follows:

Experience Criteria

Bidders were required to show experience in similar nature of work during the last three financial years. The minimum project experience requirement included either:

Proof of work had to include work orders, engagement letters, completion certificates, user acceptance testing certificates, letters of award or client-certified evidence.

Technical Evaluation Criteria

The tender followed a QCBS evaluation model with 70% technical weightage and 30% financial weightage. A minimum technical score of 70 marks was required to qualify for the financial bid evaluation.

Financial Evaluation

Only technically qualified bidders were considered for financial evaluation. The lowest qualifying financial bidder was to receive 100% financial score, and other bidders were to be scored proportionately using the prescribed formula. The final composite bid score was calculated using 70% technical score and 30% financial score.

A nil or conditional financial quote was to be treated as unresponsive. The financial bid format included cost heads for development and deployment, cloud hosting infrastructure, storage charges, and annual maintenance cost.

Tender Fee and EMD

No tender fee was required. However, bidders were required to submit an Earnest Money Deposit of Rs. 2.50 lakh.

The EMD had to be submitted in the form of Demand Draft, Banker’s Cheque or Bank Guarantee only, drawn in favour of “Pay and Accounts Officer, Ministry of Culture, New Delhi”, payable at New Delhi. The EMD was required to remain valid for 90 days.

EMD exemption was available for eligible MSE/MSME/NSIC/SSI bidders, subject to submission of valid proof and self-declaration.

Submission Guidelines

The tender was to be submitted online through the Central Public Procurement Portal. Bidders were required to use valid Digital Signature Certificates for online submission.

The bid submission followed a two-envelope system:

1. Technical Envelope: Included eligibility documents, scanned EMD instrument or exemption proof, registration documents, work experience documents, financial eligibility documents, certifications and declarations.
2. Financial Envelope: Included the rate quote sheet in XLS format, duly completed as per the prescribed BOQ / financial bid format.

The physical EMD instrument was required to be submitted before the opening of the technical bid, wherever applicable.

Document Checklist

Bidders were required to prepare and submit the following documents:

Final Result

As per the corrigendum dated 08.06.2026, seven bidders qualified for financial bid evaluation. After assessment of technical presentation scores and financial bids, Tagbin Services Private Ltd. secured the highest composite score of 84.022 and was ranked H1. The Ministry of Culture awarded the work for development and deployment of AI integration on the Ministry of Culture website and subsequent maintenance and support to Tagbin Services Private Ltd.

Bidder Ranking