Learn from Competition, Apply through Learning | 1981 Tech AI Contest

A month of research, study and competition has integrated AI into daily operations

Preparation period: August 1 – August 31, 2026

The first AI Skill Application Competition of the Operations Department has concluded successfully. Participating teams selected topics independently based on real business scenarios. After one month of research and refinement, the top three winners were finally determined through votes from all staff.

Turn AI from something "heard of" into something "usable"

Cross-border e-commerce features a long and fragmented operation chain, including product selection, copywriting, graphic design, video production, advertising, data analysis and inventory preparation. Human resources are consumed at every stage. The starting point of this competition is to move AI beyond conceptual discussion and test it in real business scenarios.

The competition focused on pre-match preparation: each team identified their respective pain points, learned while applying knowledge and made improvements in practice. In the end, representatives were selected to clearly present "what problems have been solved and how they were addressed". All participants scored the teams from two dimensions: "learning gains" and "satisfaction level".

Assess the market carefully before selecting products.

Some teams are handing market research over to AI: competitor structures, price ranges and selling point distributions are automatically compiled. Tasks that once took three hours are cut down to 30 minutes. Product selection no longer relies on intuition, and every decision is backed by evidence.

                                                Capability module from market research to product selection

The copywriting has a unified "writing style".

Fill in the product information and submit it to AI, and it will generate core keyword titles that comply with both platform algorithms and regulatory requirements, together with bullet point descriptions and search keywords. After writing, you can use Skill for quality inspection: input a link, and a graphic inspection report will be generated within minutes, with revision suggestions provided as well.

Listing Quality Inspection and Copywriting Generation

A second pair of eyes for ad adjustments

Hand over the ad backend data to AI, and you will get actionable optimization recommendations in return. There is solid evidence for which keywords to increase bids on, which ones to add as negatives, and where to reallocate the budget. Daily monitoring shifts from "looking at numbers" to "reviewing conclusions."

A data pipeline that can fetch data automatically

Some teams also focus on "enabling AI to retrieve data on its own": there is no need to write code or remember tool names. As long as the problem is clearly described, the AI will automatically call appropriate data tools and complete the full-link analysis in a single conversation.

Let AI automatically call data tools

Comparison of Efficiency Improvement in 7 Major Scenarios

A production line for chart making and video production

Content-side changes alter the workflow: first use AI to plan the graphic design scheme, predefine the selling points, copywriting and background color for each main image, and then start image creation. This enables stable batch production of main images and A+ pages; product images can also be directly used to generate short video materials.

Main image plan designed by AI

A dedicated operation workstation of your own

Some teams also use AI to build workbenches tailored to their own businesses, bringing pending tasks for multiple stores, managed products and cost structures onto a single page. The supporting knowledge base is divided into 10 modules including market research, Listings, advertising, data analysis and after-sales service, with speech templates available for one-click import and export.

Self-built AI operation workstation

A replenishment model with recalculation capability

Inventory age, available stock, in-transit goods, sales volume and procurement lead time are consolidated into one table, turning "whether to replenish stock and how much to replenish" into a recalculable formula.

Honor Roll

After review and scoring by all participants, Operation Department 5 won the first place, Operation Department 7 took the second place, and Operation Department 9 got the third place.

Next step

These methods, templates and tools will be continuously shared within the department and gradually rolled out to more sites and teams, so that AI is not merely a highlight during competitions, but the most handy tool for daily work.

Scan the QR code to consult the 1981 team if you need to learn more details about cross-border e-commerce operations.