Expert Services (human team)
Our automated workflow compresses the literature work as far as it can. But some things a machine cannot do — and should not do: deciding whether an idea is worth pursuing, getting a sample-size estimate to hold up, interpreting non-significant data, or rewriting a manuscript so it reads like it was written by someone in the field. Those we do ourselves.
Four services
Human polishing
- Sentence rhythm, terminology consistency, and tense and tone calibration in Chinese or English manuscripts
- Paragraph logic and gaps in the argument flagged for you (we do not rewrite your conclusions — we point out where they do not hold)
- Citation verification before delivery: whether each source really supports the claim it is attached to
- A verification sheet accompanies the manuscript (citation traceability, formatting check, hidden-character and homoglyph detection)
Expert consultation
- A 30-minute video or voice call, working through your question list item by item
- Direction calls: whether an idea is publishable, and at what tier of journal to aim
- You get judgement and the reasoning behind it — we do not decide for you
Study design
- Hypothesis and endpoint definitions, groups and controls, inclusion and exclusion criteria
- Inputs and assumptions for sample-size estimation (reproducible calculation, no black box)
- Confounder control and pre-specified statistical methods (so methods are not chosen after seeing results)
- A checklist of material needed ahead of ethics and statistics approval
Troubleshooting consultation
- Working through failures item by item: no results, poor reproducibility, models that will not converge
- How to present and interpret non-significant or opposite-direction results honestly
- Whether the data can be used, whether to add experiments, and how many
Service boundaries (please read first)
What we do:Polishing at the language and logic level, methodological judgement, design advice, and troubleshooting.
What we do not do (we will decline these):
- No ghostwriting: we do not write papers or grant applications for you, and do not form scientific judgements or conclusions in your place
- No plagiarism-evasion rewriting: we do not provide rewrites aimed at evading similarity checks or concealing the origin of the writing
- No fabricated or cosmeticised data: we do not produce data, adjust statistical results, or create figures and citations that do not exist
- No acceptance promises: we make no commitment about journal acceptance or review outcomes
Consultation does not replace your institution's ethics review, statistics approval or author responsibility; the manuscript content and authorship remain yours. Disclosing AI use as your target journal requires is something we expect of partners, not an optional extra.
Pricing
On request: quoted individually by manuscript length, specialty and turnaround. We do not publish list prices — it avoids arbitrary comparison and lets us price the actual job fairly.
Contact
Email: supports@zhiyue-ai.cn(WeChat/phone provided in our reply after you submit a request)
Please include: word count or experiment type, specialty, desired turnaround, and the services you would accept. We reply within 24 hours.