Knowledge Graph

The end-to-end chain from sourcing and warehousing to fulfillment — the hub of all my methodologies.

The 30-minute delivery business: instant demand, instant fulfillment — and decisions must be instant too.

Small warehouses closest to users: compact, fast-moving, spoilage-sensitive — the training ground for inventory strategy.

Walking the line between stockouts and overstock: forecast-driven, role-based water levels, real-time correction.

A function of service level and demand volatility. The static formula is simple; dynamic implementation is hard.

Data governance > feature engineering > evaluation > model choice — in that order.

Traffic / profit / image / convenience — ranking SKUs is the master switch of differentiated strategy.

The trade-off between price perception and margin — KVIs decide whether users think you're expensive.

The few items whose prices users actually remember — benchmark competitors frequently and never yield.

What to stock in limited space: balancing traffic, margin and perception structure.

The core capital-efficiency metric — push too low and you pay in stockouts.

The churn switch of instant retail: one stockout = one reason to uninstall.

The invisible margin killer — at the intersection of turnover, forecasting and replenishment cadence.

Order-to-door speed, jointly determined by network layout and wave design.

Where and how big to build: center-of-gravity, coverage models and simulation rehearsals.

Rehearse real decisions in a virtual environment, driving trial-and-error cost to near zero.

Making algorithms serve business goals: metric definition, strategy frameworks and rollout cadence.

Computer vision for supermarket loss prevention and camera patrol — scenario definition matters more than model accuracy.

The other end of the retail chain: tracking, A/B testing and user tags — the groundwork of data-driven growth.

SQL / Python / statistics / visualization — the hands-on foundation of every methodology.