Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.01119 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911413853749248 |
|---|---|
| author | Chernyshev, Konstantin Artemova, Ekaterina Zhukov, Viacheslav Nerush, Maksim Fedorova, Mariia Repik, Iryna Shapovalova, Olga Sukhorosov, Aleksey Dobrovolskii, Vladimir Mikhailova, Natalia Tilga, Sergei |
| author_facet | Chernyshev, Konstantin Artemova, Ekaterina Zhukov, Viacheslav Nerush, Maksim Fedorova, Mariia Repik, Iryna Shapovalova, Olga Sukhorosov, Aleksey Dobrovolskii, Vladimir Mikhailova, Natalia Tilga, Sergei |
| contents | Tendem is a hybrid system where AI handles structured, repeatable work and Human Experts step in when the models fail or to verify results. Each result undergoes a comprehensive quality review before delivery to the Client. To assess Tendem's performance, we conducted a series of in-house evaluations on 94 real-world tasks, comparing it with AI-only agents and human-only workflows carried out by Upwork freelancers. The results show that Tendem consistently delivers higher-quality outputs with faster turnaround times. At the same time, its operational costs remain comparable to human-only execution. On third-party agentic benchmarks, Tendem's AI Agent (operating autonomously, without human involvement) performs near state-of-the-art on web browsing and tool-use tasks while demonstrating strong results in frontier domain knowledge and reasoning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_01119 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Tendem: A Hybrid AI+Human Platform Chernyshev, Konstantin Artemova, Ekaterina Zhukov, Viacheslav Nerush, Maksim Fedorova, Mariia Repik, Iryna Shapovalova, Olga Sukhorosov, Aleksey Dobrovolskii, Vladimir Mikhailova, Natalia Tilga, Sergei Computation and Language Tendem is a hybrid system where AI handles structured, repeatable work and Human Experts step in when the models fail or to verify results. Each result undergoes a comprehensive quality review before delivery to the Client. To assess Tendem's performance, we conducted a series of in-house evaluations on 94 real-world tasks, comparing it with AI-only agents and human-only workflows carried out by Upwork freelancers. The results show that Tendem consistently delivers higher-quality outputs with faster turnaround times. At the same time, its operational costs remain comparable to human-only execution. On third-party agentic benchmarks, Tendem's AI Agent (operating autonomously, without human involvement) performs near state-of-the-art on web browsing and tool-use tasks while demonstrating strong results in frontier domain knowledge and reasoning. |
| title | Tendem: A Hybrid AI+Human Platform |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2602.01119 |