Method

Reinforcement learning

Value or policy learning (DQN, DDPG) incl. offline RL.

21 cases from 17 companies apply Reinforcement learning. 7 were checked in an online A/B test; 2 report results in production. Includes variants: PPO.

Typical tasks: Optimize decisions, Build an assistant over internal docs, Rank search results, Recommend from what the user just did and 7 more. Companies: Netflix, Yandex, Rabota.ru, Databricks, Sber, Wix and 11 more. Often combined with: Reward model, PPO, Supervised fine-tuning, Multi-armed bandit, Synthetic data generation (LLM) and 66 more.

How XP Drove a 72% Lift in Account Opens and 35% Activation Increase with Agentic CRMfintech2026-06
Thinking Fast & Slow: How Databricks Built High-Speed and Deep Research Agentsgeneral2026-06
How a 4B Model Beat GPT at Agentic Retrieval (And How to Train One on a Neon Postgres Corpus)general2026-06
ML in Store Operations and Logistics: How Production Models Directly Optimize Processesretail2026-05
The Search of the Futuregeneral2026-05
Managing Oil and Gas Field Development with a Multi-Agent System and Reinforcement Learningmanufacturing2026-05
Post-Training Generative Recommenders with Advantage-Weighted Supervised Finetuningmedia-streaming2025-10
Product Applications of Yandex VLMsearch-engine2025-05
AI for Revolutionizing Customer Care Routing System at Wixsaas-b2b2024-08
Advanced Use of GigaChat at Rabota.ru: Synthetic Data and Searchclassifieds2024-05
Reading Between the Lines: How to Use ML to Curate Personalized Collections Better Than Editorsfintech2024-05
Technology Trends in RecSysgeneral2024-05
Well Placement for Drilling Using Reinforcement Learningmanufacturing2024-05
Using Contextual Bandit models in large action spaces at Instacarte-commerce2023-06
A Job Posting Summary Generation Service Based on an LM Fine-Tuned with RLHFclassifieds2023-05
A Routing Algorithm Without a Trade-off Between Quality and Speedfood-delivery2023-05
RLHF for retrieval based NLP systemgeneral2023-05
LLaMA and a Zillion Ways to Fine-Tune It—general2023-05
A Gallop Through Production Sites: ML at Nornickelmanufacturing2023-05
Reinforcement Learning for Budget Constrained Recommendationsmedia-streaming2022-08
Pricing at Lyftmobility2022-08