Case Study: ShareChat accelerates ML issue detection and boosts engagement and inclusivity with Arize AI

ShareChat’s Machine Learning Team Grows Engagement, Inclusivity

ShareChat, a social media unicorn with over 400 million monthly active users across ShareChat, Moj, and MX TakaTak, depends on machine learning for advertising, feed ranking, and content moderation. The ML organization faced slow detection of performance issues (about 24 hours), time-consuming ad‑hoc troubleshooting across hundreds of models, and limited tooling for monitoring unstructured data—so ShareChat selected Arize AI and its Arize for ML observability platform as their ML observability partner.

Arize AI’s platform delivered pre‑launch validation, automated monitors, structured and unstructured data monitoring, drift and bias tracing, and end‑to‑end performance tracing, enabling ShareChat to detect problems in real time and pinpoint root causes. As a result, the team freed up hundreds of hours per year, achieved a payback period under one year (>100% ROI), moved from 24‑hour lagging dashboards to immediate alerts, improved model performance and fairness monitoring, and reduced mean time‑to‑resolution so engineers can focus on core ML work.

ShareChat

Sourav Maitra

Technical Lead, AI - Ad Relevance