TareProgram

What it is

TareProgram is a confidential post-processing service for organizations that rely on long-baseline GNSS clock comparison (final clock products and related comparison streams). You submit data in accepted pilot formats; we return corrected outputs and statistical comparisons for your internal evaluation. We do not disclose how the correction is formed—no equations, parameters, source code, or implementation details. Processing is patent pending. Your data remains yours. Effect sizes and validation design (still no method) are in the technical note.

Evaluation frame

Definitions only—no mechanism. We state clearly what is compared: the same agreed primary comparison/residual series and the same metric (principally RMS) before vs after our confidential post-process; reported % improvement is that statistic's relative gain, with holdout meaning no refit on the evaluation window.

What you receive (form only): Evaluation outputs—typically time-aligned corrected traces (series/residuals) in agreed formats; summary statistics (RMS and anything else named in your agreement); where specified, consistency- or deviation-style indicators; and optional brief review-level signals—not a separate calibrated uncertainty product unless your pack explicitly says so. No description of how outputs are produced.

Details match the pilot one-pager and technical note.

Who it's for

National metrology institutes, UTC contributors, fundamental physics experiments (e.g. LPI, variation of constants), and anyone who cares about sub-nanosecond-level clock comparison over long baselines. We are building a service with clear rules, fixed schedules, and rigorous data handling.

Technical note

Results and validation summary (no method disclosure): effect size, validation protocol, and blind pilot proposal. Two-page summary; no form, parameters, or implementation disclosed.

Pair with the demonstration: the note carries audited numbers, windows, and holdout design; the demo is a schematic walkthrough in the same IGS clock-comparison setting—not a substitute for this document.

Demonstration

Schematic before/after visuals grounded in the same public IGS clock-comparison domain as the technical note—figures are illustrative, not a dump of raw run artifacts. No method or implementation disclosed.

Pair with the technical note: use this page for shape and intuition; use the technical note for what was validated and how.

Pilot one-pager

Proprietary timing integrity, results-only. GNSS clock comparison: typical held-out RMS reduction around ~45%, strict-window runs up to ~81%, post-correction residuals often in the 17–22 ns range (see technical note). Plus an illustrative event-ordering intuition (50 ns separation; 99% vs 64% between two uncertainty regimes)—not a second field claim. For labs, HFT desks, telecom, and PPP teams.

Built for pilot-to-deployment decisions: same dataset, same metric, auditable before/after outcomes under a holdout protocol.

FAQ and pilot terms

Clear answers on what Tare is, who can participate, data handling, cost, and IP. Pilot terms govern the 90-day program.