AI won't fully replace analysts: Avito's takeaways

On September 30, 2026, AvitoTech published a 16-minute video, “Will AI replace analysts?”, in which three Avito analysts explain why neural networks won't take their jobs entirely: analysis results are too generic, hard skills are becoming cheaper, while domain expertise and the ability to work with AI agents are becoming more valuable.

Compiled by Sergey KostenchukPublished 2026-10-01Updated 2026-10-01

2026-10-01 Coding

💻 AI won't fully replace analysts

AvitoTech published a 16-minute video, “Will AI replace analysts?”, on September 30, 2026, featuring three of its analysts. Their conclusion: the model produces results that are too generic, and a significant portion of the work has to be redone manually.

🌍 Routine SQL and code are delegated to LLMs, while value shifts to problem formulation, context gathering, and result validation. The analyst-AI tandem is already in production at Avito: the corporate assistant “Vitalik” operates on an LLM portal and an agent factory.

👤 It is more useful for analysts to learn how to ask questions and check LLM answers for bias than simply to write queries faster. The video features three cases: migrating 200 data marts in six months, automated checking of employee work, and training an LLM to analyze metrics — but without a baseline and metrics, these are practitioners' impressions.

Source 1: https://www.youtube.com/watch?v=RBChn_YTcWs Source 2: https://habr.com/ru/companies/avito/articles/1071126/

Sources