International Journal of Innovative Research in Computer and Communication Engineering

ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines

| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |


TITLE Multi-Tenant Retrieval-Augmented Generation Using a Shared Pinecone Vector Index
ABSTRACT Deploying retrieval-augmented generation (RAG) as a multi-tenant product requires durable vector storage, tenant isolation, and a simple embedding path. This paper presents a Pinecone-centred architecture in which all chatbots share one serverless index. Isolation is enforced by tagging every vector with a chatbot identifier and applying an equality metadata filter on every query, while Pinecone hosted inference (multilingual-e5-large) generates embeddings. The design is realised in Relay, a Next.js platform with a dashboard and an embeddable widget. A functional evaluation on a deployed instance supports it as a blueprint needing no per-tenant indexes or Python embedding sidecar.
AUTHOR AADITI CHORAMLE, CHITRANG CHOUDHARI, PRATIK KAMBLE, SAHIL DHANAVADE, DR. SWATI D. GHULE MCA Student, Department of Master of Computer Application, P.E.S.’s Modern College of Engineering, Pune, Maharashtra, India Professor and Project Guide, Department of Master of Computer Application, P.E.S.’s Modern College of Engineering, Pune, Maharashtra, India
VOLUME 189
DOI DOI: 10.15680/IJIRCCE.2026.1410037
PDF pdf/37_Multi-Tenant Retrieval-Augmented Generation Using a Shared Pinecone Vector Index.pdf
KEYWORDS
References [1] A. Vaswani et al., “Attention is all you need,” in Proc. NeurIPS, 2017, pp. 5998–6008.
[2] Z. Ji et al., “Survey of hallucination in natural language generation,” ACM Comput. Surv., vol. 55, no. 12, pp. 1–38, 2023.
[3] P. Lewis et al., “Retrieval-augmented generation for knowledge-intensive NLP tasks,” in Proc. NeurIPS, vol. 33, 2020, pp. 9459–9474.
[4] V. Karpukhin et al., “Dense passage retrieval for open-domain question answering,” in Proc. EMNLP, 2020, pp. 6769–6781.
[5] N. Reimers and I. Gurevych, “Sentence-BERT: Sentence embeddings using Siamese BERT-networks,” in Proc. EMNLP-IJCNLP, 2019, pp. 3982–3992.
[6] J. Johnson, M. Douze, and H. Jégou, “Billion-scale similarity search with GPUs,” IEEE Trans. Big Data, vol. 7, no. 3, pp. 535–547, 2019.
[7] J. J. Pan, J. Wang, and G. Li, “Survey of vector database management systems,” VLDB J., vol. 33, pp. 1591–1615, 2024.
[8] Pinecone Systems, “Pinecone documentation: Indexes, metadata filtering, and inference,” 2024. [Online]. Available: https://docs.pinecone.io
[9] G. Izacard and E. Grave, “Leveraging passage retrieval with generative models for open domain question answering,” in Proc. EACL, 2021, pp. 874–880.
[10] Y. Gao et al., “Retrieval-augmented generation for large language models: A survey,” arXiv:2312.10997, 2023.
[11] A. Asai, Z. Wu, Y. Wang, A. Sil, and H. Hajishirzi, “Self-RAG: Learning to retrieve, generate, and critique through self-reflection,” in Proc. ICLR, 2024.
[12] L. Wang et al., “Multilingual E5 text embeddings: A technical report,” arXiv:2402.05672, 2024.
[13] R. Krebs, C. Momm, and S. Kounev, “Architectural concerns in multi-tenant SaaS applications,” in Proc. CLOSER, 2012, pp. 426–431.
[14] H. Chase, “LangChain documentation,” 2022. [Online]. Available: https://python.langchain.com
[15] Auth.js, “Auth.js documentation (NextAuth.js v5).” [Online]. Available: https://authjs.dev
[16] MongoDB Inc., “MongoDB manual.” [Online]. Available: https://www.mongodb.com/docs/manual/
[17] OpenRouter, “OpenRouter API reference.” [Online]. Available: https://openrouter.ai/docs
[18] OWASP Foundation, “OWASP Top 10 – 2021.” [Online]. Available: https://owasp.org/Top10/
[19] Vercel Inc., “Next.js documentation.” [Online]. Available: https://nextjs.org/docs
[20] A. Grattafiori et al., “The Llama 3 herd of models,” arXiv:2407.21783, 2024.


Copyright © IJIRCCE 2020.All right reserved