Artificial Intelligence 15 min read

Xiaomi’s Experience in Deploying Intelligent Analytics: Productization, Challenges, and Future Plans

The article shares Xiaomi’s practical experience in building and productizing intelligent analytics, explaining why it is needed, how it integrates with BI, the essential prerequisites, staged implementation, technical challenges, and future roadmap including smart alerts, automated insights, and data Q&A.

DataFunSummit
DataFunSummit
DataFunSummit
Xiaomi’s Experience in Deploying Intelligent Analytics: Productization, Challenges, and Future Plans

Why Intelligent Analytics? Xiaomi identifies a growing need for intelligent analytics due to increasing data‑driven decision making, analyst shortages, high manual effort in anomaly detection, and the desire for faster, more agile business insights.

Combining Intelligent Analytics with BI The team argues that embedding analytics as a plug‑in to existing BI reports offers a familiar user experience, enables scalable productization, and avoids the fragility of isolated algorithm packages.

Industry Productization Issues Existing BI tools like Tableau and Power BI are seen as too generic and introduce extra learning curves; Xiaomi therefore focuses on reducing user friction by adding incremental value within the current analysis workflow.

Productization Practice

1. Prerequisites : a recent overhaul of the BI platform’s data modeling and query services, high report usage, and a mature algorithm library.

2. Three Stages : (a) automate routine analyst tasks, (b) assist analysts with complex or large‑scale calculations, (c) enable fully autonomous insight generation.

3. Finding the Minimum Viable Product : partner with a data‑intensive advertising business that has frequent anomaly‑analysis needs, starting with anomaly detection and root‑cause analysis on reports.

4. Implementation Challenges

• Translating business problems into generic algorithms and handling complex metric formulas.

• Query performance for large data volumes, leading to a hybrid real‑time/offline approach and pre‑computation for high‑frequency scenarios.

• Varying data schemas across businesses, requiring data lineage capabilities to trace root causes beyond aggregated tables.

Future Planning

Current capability building includes smart alerts, automated story generation, data‑question answering (NLP), and a data encyclopedia that stores metric metadata and lineage.

The vision is to shift from “people looking for data” to “data finding people,” delivering insights via push notifications, IM cards, and integrated workflow tools.

Finally, the speaker thanks the audience for their attention.

big dataAIdata platformproductizationBI Integrationintelligent analytics
DataFunSummit
Written by

DataFunSummit

Official account of the DataFun community, dedicated to sharing big data and AI industry summit news and speaker talks, with regular downloadable resource packs.

0 followers
Reader feedback

How this landed with the community

login Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.