An open institute and a co-learning cooperative. We work on questions that have not yet matured — and turn them, step by step, into research, tools, tutorials, and specs. Everything is built in the open, and every contributor keeps their name on the work.
Our process has seven steps: a question is raised, defined, opened for collaborators, built together, reviewed, delivered, and archived. The community is nonprofit and open-source first; we do not run ads, and we do not sell access.
Featured research: the FDE series
Forward deployed engineers became the most-hired role in AI within a year. We read the primary sources — the job posts, the Palantir retrospectives, the practitioner threads — and wrote a three-part series in English. All three parts are out now:
- Part 1 — Why Anthropic and OpenAI bet billions on the same twenty-year-old job
What an FDE is, why it is happening now, and why most imitators will fail. - Part 2 — China's FDE hiring boom: same name, different constraints
The U.S. FDE was summoned by models too new to deploy themselves. The Chinese FDE is pinned in place by data that legally cannot leave the building. - Part 3 — The FDE career ledger: founder training, not an engineering ladder
Pay, daily reality, career paths, and three questions to spot a fake FDE role.
Everything else is in Chinese — for now
Our main site is Chinese-language: 20 research articles, open research issues, a paper-search tool, and community projects. If you read Chinese, start from the homepage. English editions of future research will land here, at /en/.
Contact
- Email: kezhongke_htac@163.com
- GitHub: KezhongKe-poieticlab