This site brings together Anant Mishra's technology writing, architectural work, and developing research programme. Each format serves a different purpose. A useful idea should be easy to understand, and its evidence should be easy to inspect.
What each label means
Essays present analysis and recommendations. Design studies examine a proposed architecture or an explicitly identified simulation. Case studies describe work and outcomes, with the source and limits of reported measurements stated. Research protocols set out questions and planned methods before results are available. Existing architecture frameworks describe engineering approaches and do not establish peer-reviewed findings.
Sources and claims
New articles link to relevant primary sources where possible. A source supports the claim beside it; it does not establish that a proposed system has been deployed or independently validated. Quotations require a verifiable source or the speaker's approval.
Performance and outcome claims need a defined measurement period, baseline, method, and evidence. Where client material is confidential, that limitation should be stated. Existing portfolio figures are author-reported unless a page links independent verification.
Portfolio evidence
The $100M+ pipeline, 200+ people led, and two exits are author-reported career figures. Pipeline is not recognised revenue. Supporting company records, measurement definitions and time periods are not published on this site.
Older case studies and essays also report figures such as 65.4% inference-cost savings, 99.98% uptime and a $4.2M historical inventory discrepancy. The underlying audit records and datasets are not publicly available here. These figures should not be read as independently verified results, guarantees, or outcomes established by the newer synthetic studies.
Executive roles and affiliations
Current IAG and Invadecore leadership roles, the Namibia and Dubai CEO mandate and the recent acquisition description are executive-provided. The personal site does not reproduce independently verifiable corporate appointment or transaction records. Company records and official announcements should be used for formal diligence.
Synthetic work and research
Synthetic scenarios must identify their assumptions and intended use. They can support engineering tests and policy comparisons within a model. Claims of real-world effectiveness require appropriate field evidence.
Research pages state their stage. A protocol is not a completed paper. Publication records, peer review, identifiers, affiliations, and scholarly indexing are added only after they can be verified. The consistent research byline is Anant Mishra; another author's work must never be attributed on the basis of a shared name.
Review and corrections
Routine website publication is authorised by the author and follows editorial, source, and evidence checks. Research manuscripts require accountable human authorship and verification, with any AI-assistance disclosure required by the destination publication. A substantive update should record its actual revision date. Rebuilding the website does not create a new publication date.
For a correction, contact Anant with the page URL, the claim in question, and the supporting evidence. Corrections should be reflected in both visible content and associated metadata.
Site measurement
When enabled, Vercel Web Analytics measures page views and clicks on email and booking links. Contact-click events contain the public page path, not email contents, recipient addresses or booking details. A click records interest, not a completed enquiry. Query values and URL fragments are excluded from the site's event URLs.