CLIENT CASES / Client Cases
Solve one real problem first, then talk about scale
The six cases below sit in intelligent manufacturing, higher & vocational education, and cross-border compliance. The approaches differ but the thinking is the same: start from the most painful scenario, validate with real data, and expand only once it works.
01 / ANONYMISED CASES
Six delivered scenarios
01
A large electrical equipment group · Equipment maintenance knowledge base and intelligent Q&A
- Context
- Equipment manuals, process documents and past work orders were spread across several business systems. Frontline staff relied on phone calls and personal experience, and the same questions came up again and again.
- Approach
- Multi-format documents were parsed and ingested into a unified enterprise knowledge base with retrieval-based Q&A; authorisation was tiered by business unit and role; the whole system was deployed inside the company's own cloud, with data staying in region.
Outcome
- Document search went from 20–30 minutes on average to under one minute
- Answers cite their source text and can be reviewed and traced
- Recurring questions accumulate in the knowledge base, cutting repeat enquiries
02
A vocational college · Teaching Q&A and practical-training guide assistant
- Context
- Course knowledge was scattered across slides, practical-training manuals and individual teachers' experience, and student questions concentrated on a few teachers. Equipment procedures were hard to demonstrate repeatedly — one person explaining, many watching.
- Approach
- We built a course knowledge base and teaching Q&A assistant with authorisation tiered by programme, course and class; turned equipment procedures into step-by-step, queryable guidance; and gave teachers an aggregated view of frequent questions for course improvement.
Outcome
- Students resolve common questions themselves, reducing repeat explanations
- Procedures move from flipping through a manual to a single query
- Frequent questions become teaching assets reusable across cohorts
03
A listed manufacturer · After-sales and channel technical support assistant
- Context
- Overseas distributors and end customers asked highly repetitive questions, consuming a large share of technical-support capacity on basics, with long waits across time zones.
- Approach
- We organised the after-sales knowledge base and published it as a multi-channel assistant (web / WeCom / official account), added human handover and escalation, and fed unresolved questions back into the knowledge base automatically.
Outcome
- Frontline and overseas channel users resolve common issues themselves
- Technical support focuses on complex problems, with faster cross-time-zone response
- Coverage grows as the knowledge base is continuously extended
04
A cross-border consumer goods group · Product knowledge and multilingual content assistant
- Context
- Product information, promotion policies and operations manuals were voluminous and updated often; several markets needed different language versions and the wording was hard to keep consistent.
- Approach
- We built a product and operations knowledge base as the single source of truth, and provided content-generation assistance that drafts multilingual copy to brand and compliance guidelines, with all output going through human final review.
Outcome
- Drafting throughput improved, with people focused on review and creativity
- Consistent messaging across markets, with fewer wording discrepancies
- Policy updates propagate quickly to every channel
05
A manufacturing group · Supplier ESG data collection and attestation
- Context
- Overseas customers and regulators required carbon, energy and labour-compliance data from the supply chain. Suppliers were spread across countries and chased repeatedly by email and Excel, with inconsistent definitions that were hard to verify.
- Approach
- Supplier ESG questionnaires were moved online with automatic triage and answer-quality checks; metric definitions were unified and outliers cross-checked; key data was attested on-chain to form a traceable data chain.
Outcome
- Questionnaire collection and checking shifted from manual chasing to system-driven
- Data provenance and change history are traceable and tamper-resistant
- Disclosure working papers generated for CSRD and similar requirements, cutting last-minute scrambling
06
An export consumer-goods company · PPWR packaging compliance readiness
- Context
- The EU PPWR has started to apply, and customers and channels are asking for evidence on packaging material, recyclability and recycled content. The company's data was spread across departments and nobody could give a complete answer.
- Approach
- We established a packaging list and component BOM, compared recyclability, recycled content, weight reduction and empty-space requirements item by item, and produced a gap list with remediation priorities, alongside DoC technical files and EPR material for each market.
Outcome
- Packaging data is centrally registered and traceable to material and supplier
- Gaps and remediation priorities are clear rather than guessed
- Requests for evidence can be answered immediately from the register
About these cases
The cases above have been anonymised: no client names or commercial secrets are disclosed, and the focus is on the scenario and the approach. Actual results and timelines depend on data quality, system conditions and the degree of cooperation, and are governed by the project proposal and acceptance results.
The cases above have been anonymised: no client names or commercial secrets are disclosed, and the focus is on the scenario and the approach. Actual results and timelines depend on data quality, system conditions and the degree of cooperation, and are governed by the project proposal and acceptance results.
02 / CLIENTS
Some of the enterprise clients we have served
Spanning manufacturing, energy, retail, consumer goods and professional services — many of them expanding overseas.
Shanghai Electric
BYD
GAC Group
Dongfeng Commercial Vehicle
Foton Cummins
Wanxiang Group
Foxconn
Sany Heavy Industry
Shanghai EMU
CISDI
Morimatsu
Emerson
BBCA Group
PetroChina
SPIC
China Mobile
21Vianet
China Aviation Oil
GCL Power
Unilever
L'Oréal
Coca-Cola
Budweiser
Henkel
Swire Properties
Jahwa
Shanghai Tobacco
Microsoft
Samsung
Audi
Canon
ZEISS
EY
Nokia Bell Labs
Clobotics
SPD Bank
Ping An
Roche
Merck
Bayer
Medtronic
The above are some of the enterprise clients our team has served, in no particular order; the exact project scope and content are as agreed in contract. If you would like details relevant to your industry, please email us.
03 / NEXT STEP
Want to know how this works in your industry?
Tell us the industry and the scenario, and we will set out the relevant project experience and the pitfalls along with it.