Open Telco release improves telecom AI models and datasets worldwide
OTel: Open Telco AI Datasets, Benchmarks, and Models
Artificial Intelligence
Summary
Telecom companies have lots of data that can help improve AI models for understanding and organizing information. The authors created Open Telco (OTel), a shared collection of telecom datasets and ready-to-use AI models for tasks like searching and ranking information. They tested these models and showed that training them on OTel data makes them work better. Many people are already using these models and datasets to build smarter telecom AI systems.
What this means in practice
- •For telecom ai engineers: Train and test telecom-specific AI models for faster and more accurate information retrieval and decision making in telecom systems.
- •For customer service ai developers: Improve customer support chatbots by using context-grounded telecom language models trained on OTel data to handle telecom queries more accurately.
Authors
Farbod Tavakkoli, Gregory Diamos, Kenneth Church, David Kanter, Mark Austin, Imtiaz Karim, Mirza Masfiqur Rahman, Merouane Abdelkader Debbah, Zeinab Nezami, Ali Maatouk, Leandros Tassiulas, Rex Ying, Nick Sorros, Louis Powell, Nikolaos Vasiloglou, Ashish Vaswani, Somanshu Singla, Adarsh Chaluvaraju
Abstract
We present Open Telco (OTel), an open telecom AI resource that releases derived telecom datasets for retrieval, reranking, instruction tuning, and safety/abstention, together with 30 full-parameter post-trained baselines spanning 10 embedding models, 3 rerankers, and 17 language models. The community has already engaged substantially with the resource: as of May 3, 2026, the released models have been downloaded over 16 million times and the project has received 157+ pieces of media coverage worldwide. Building on prior open telecom datasets and benchmarks, OTel provides documented telecom data sources, held-out evaluation partitions, trained embedding models, rerankers, context-grounded LLMs, and safety/abstention data in one unified resource. Each baseline starts from an open-weight model and is post-trained on OTel-derived data using an open training recipe, then evaluated on held-out OTel evaluation partitions. OTel post-training improves performance across all three model families: embedding retrieval reaches 93.1% NDCG@10, reranking reaches 0.947 MRR@10, and language-model correctness reaches 87.8%. We release OTel as a reproducible starting point and invite the community to expand the data, improve embedding and reranking models, and build stronger context-grounded telecom LLMs.