Trust is part of the product architecture.

1. Lawful and authorised sourcing

Sericant intends to use appropriately licensed, authorised or otherwise lawfully usable company information.

Production data integrations are intended to be activated only where relevant data rights, contractual permissions, permitted-use terms and processing arrangements are in place.

Technical accessibility to information is not treated as equivalent to having the legal right to use, process or transfer that information.

2. Provenance and timestamps

Sericant's product roadmap includes retaining source provenance, update timestamps and other relevant metadata alongside structured facts used in research outputs.

The objective is to make company intelligence more traceable and reviewable, rather than presenting AI-generated text as an unsupported factual conclusion.

3. Source facts and AI interpretation

Sericant intends to distinguish source facts from AI-generated interpretation.

AI is intended to organise, explain and synthesise approved inputs rather than invent factual corporate information. Where practical, research outputs should enable users to understand which elements originate from source information and which elements represent AI-assisted analysis or summarisation.

4. Bounded agentic research

Sericant's long-term product direction is a vertical agentic company-research workflow rather than an unrestricted autonomous system.

The intended workflow may include interpreting a research request, planning research tasks, retrieving approved-source information, resolving company entities, extracting relevant evidence, checking consistency, synthesising findings and presenting provenance.

These agentic capabilities are roadmap items unless specifically identified as already deployed.

5. Verification and human judgement

Company research can involve incomplete, conflicting or time-sensitive information. Sericant therefore intends to develop verification controls such as required-field checks, source comparison, timestamp review and missing-information detection.

Sericant is designed to assist professional research, not to replace independent human judgement in legal, compliance, investment, credit or other consequential decisions.

6. Data minimisation and purpose limitation

Sericant intends to process only information reasonably relevant to the research purpose and product function.

As production services develop, company information, personal information and higher-risk data categories are intended to be assessed separately so that unnecessary information is not collected, transferred or exposed merely because it is technically available.

7. Cross-border data governance

Sericant operates from Hong Kong for professional and cross-border company research. Any transfer or processing of regulated information across jurisdictions is assessed separately according to the relevant data category, source, contractual rights, processing purpose and applicable legal requirements.

Sericant does not assume that information may be transferred across borders merely because it is publicly accessible or technically obtainable.

8. Security and governance

As the platform progresses from demonstration to controlled production use, Sericant intends to introduce proportionate technical and organisational controls.

These may include access restrictions, audit logging, vendor review, prompt and model safeguards, data minimisation, security monitoring, incident procedures and internal review of data and AI workflows.

9. Public demo status

The interactive company-research experience currently available on this website is illustrative.

It is not a live third-party enterprise-data feed, does not imply authorisation from any specific data provider and should not be understood as evidence that Sericant has deployed a full production agentic research system.

10. Contact

Responsible AI, data and general enquiries: hello@sericant.com

Sericant Limited
Hong Kong