I’m a Research Student for Physics, not a full-time software engineer. I built PreFlight out of personal frustration with manuscript prep: checking whether cited papers have been quietly retracted, verifying claim sentences against Crossref/OpenAlex, and catching unhedged assertions before peer review.
Because research drafts often contain unpublished findings or patent-sensitive data, the core design priority was Zero Data Retention (ZDR):
- Manuscript text never leaves your browser during editing. Tokenization, AST equation isolation, and BibTeX extraction run locally in Web Workers and WASM.
- Ephemeral API calls are only made when explicitly verifying claims (querying Retraction Watch, Crossref, and OpenAlex).
- Zero manuscript persistence and zero model training.
You can test it directly in the workbench without signing up (click "Load Sample Manuscript" to see how it flags retractions): https://preflightai.tech/workbench
As someone coming from research rather than full-time software engineering, I'd really appreciate feedback from developers and academics here on:
1. How the local parser handles edge-case LaTeX environments in your papers.
2. Edge performance or UX issues you notice.
Yeah, of course we are paying for the compute on our end. I think we are waiting for our users to stress test our theory, however our secret sauce is optimizing inference so that this is a feasible product on both ends.
Like I alluded to in my initial comment, I've never understood why the inference providers of today charge per token. Whichever marketing department got our industry to be ok with being charged for the output of a REST call, I applaud.
I'm not sure it will all hold up at scale, but I am still waiting for the test to break. It helps to be a curious and optimistic person in this endeavor.
Not currently implementing any rate limits. Using an assortment of open models.
I’m a Research Student for Physics, not a full-time software engineer. I built PreFlight out of personal frustration with manuscript prep: checking whether cited papers have been quietly retracted, verifying claim sentences against Crossref/OpenAlex, and catching unhedged assertions before peer review.
Because research drafts often contain unpublished findings or patent-sensitive data, the core design priority was Zero Data Retention (ZDR): - Manuscript text never leaves your browser during editing. Tokenization, AST equation isolation, and BibTeX extraction run locally in Web Workers and WASM. - Ephemeral API calls are only made when explicitly verifying claims (querying Retraction Watch, Crossref, and OpenAlex). - Zero manuscript persistence and zero model training.
You can test it directly in the workbench without signing up (click "Load Sample Manuscript" to see how it flags retractions): https://preflightai.tech/workbench
As someone coming from research rather than full-time software engineering, I'd really appreciate feedback from developers and academics here on: 1. How the local parser handles edge-case LaTeX environments in your papers. 2. Edge performance or UX issues you notice.
Happy to answer questions!