{ "schema_version": "public-evidence-ledger/v1", "subject": { "name": "Vyoma Gajjar", "canonical_url": "https://vyomagajjar.com/" }, "verification_policy": { "checked_on": "2026-10-10", "standard": "Each record must point to a public source that names Vyoma Gajjar and directly supports the bounded claim. A search result, self-description, or generic organization page is not sufficient.", "source_classes": { "institutional-profile": "A named profile published by the institution granting the status.", "publisher-record": "A page or document published by the organization responsible for the event, course, publication, or contribution.", "public-record": "A public legal or bibliographic record naming the contributor.", "platform-record": "A public event or author page on a publishing platform. It proves the page exists, not independent endorsement of its content." }, "not_claimed": [ "No Google Scholar author profile is claimed because no unique public author profile was verified.", "These sources verify the listed role, contribution, appearance, or public record. They do not verify unrelated biography or impact claims." ] }, "project_evidence": { "repository": "https://github.com/Vyoma/agent-economics-lab", "revision": "d56e2ffa1dff77dc2985fcc68ec114da0009041c", "revision_url": "https://github.com/Vyoma/agent-economics-lab/tree/d56e2ffa1dff77dc2985fcc68ec114da0009041c", "reproduce_command": "git clone https://github.com/Vyoma/agent-economics-lab.git && git -C agent-economics-lab checkout d56e2ffa1dff77dc2985fcc68ec114da0009041c && make -C agent-economics-lab reproduce", "claim_boundary": "These are public-dataset audits, controlled software experiments, and one small paired public case. They are not model rankings, production prevalence estimates, or deployment recommendations.", "worked_decision": { "record_id": "support-assurance-demo-v1", "fixture_type": "synthetic conformance fixture", "source_url": "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/examples/agent-decision-record.json", "verifier_url": "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/agent_economics/decision_record.py", "review_url": "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/examples/agent-decision-record.html", "requested": "SCALE", "issued": "ASSIST", "required_claims": 6, "passed_claims": 2, "failed_claims": 4, "metrics": { "acceptable_rate": 0.75, "acceptable_rate_gate": 0.8, "cost_per_acceptable_outcome_usd": 3.5, "cost_cap_usd": 2.0, "p95_task_cost_usd": 14.25, "p95_task_cost_cap_usd": 8.0, "incremental_net_value_vs_baseline_usd": 2.77, "observed_max_calls_per_task": 12, "call_cap_per_task": 8 }, "evidence_digest_prefix": "6441a11a69a0", "claim_boundary": "Eight synthetic tasks. This record is not a production deployment authorization, prevalence estimate, or ROI claim." }, "claims": [ { "id": "public-datasets-audited", "claim": "Public agent datasets audited under one protocol", "value": 10, "source_url": "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/research/EVALS.md", "evidence_locator": "The instrument scorecard enumerates the ten public sources covered by the audit." }, { "id": "joined-outcome-proxy-runs", "claim": "Runs carrying both outcome and proxy signals", "value": 31389, "source_url": "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/research/FINDINGS.md", "evidence_locator": "The cross-corpus findings report gives the shared outcome-and-proxy population used by the audit." }, { "id": "controlled-gate-perturbations", "claim": "Controlled required-gate perturbations", "value": 588, "source_url": "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/research/EVALS.md", "evidence_locator": "The instrument scorecard reports the controlled perturbation count and its scope." } ], "artifacts": [ "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/templates/agent-decision-record.template.json", "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/examples/agent-decision-record.json", "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/examples/agent-decision-record.html", "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/agent_economics/decision_record.py", "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/templates/production-feedback-contract.template.json", "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/examples/production-feedback-contract.json", "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/agent_economics/feedback_contract.py", "https://github.com/Vyoma/agent-economics-lab/blob/d56e2ffa1dff77dc2985fcc68ec114da0009041c/docs/production-feedback-loops.md" ] }, "records": [ { "id": "ioa-fellow", "claim": "Vyoma Gajjar has a Fellow profile in the Institute of Analytics directory.", "source_url": "https://ioaglobal.org/ioa-fellows/vyoma-rajeshkumar-gajjar/", "publisher": "Institute of Analytics", "source_class": "institutional-profile", "evidence_locator": "The page is headed Fellow Profile and names Vyoma Rajeshkumar Gajjar.", "checked_on": "2026-10-10" }, { "id": "webby-judge", "claim": "Vyoma Gajjar served as a Webby Awards judge and Associate Member of the Academy.", "source_url": "https://www.webbyawards.com/window-into-the-academy-10-judges/", "publisher": "The Webby Awards", "source_class": "publisher-record", "evidence_locator": "The Webby Awards article lists Vyoma Gajjar as the second featured judge and describes her contribution to the Academy.", "checked_on": "2026-10-10" }, { "id": "ieee-wie-expert", "claim": "Vyoma Gajjar is listed in the IEEE Women in Engineering Industry Experts Network.", "source_url": "https://wie.ieee.org/industry/industry-experts-network/", "publisher": "IEEE Women in Engineering", "source_class": "institutional-profile", "evidence_locator": "The directory names Vyoma Gajjar, identifies her AI expertise, and provides her talk topic and biography.", "checked_on": "2026-10-10" }, { "id": "us-patent-11748453", "claim": "Vyoma Gajjar is a named inventor on US Patent 11,748,453.", "source_url": "https://patents.google.com/patent/US11748453B2/en", "publisher": "Google Patents", "source_class": "public-record", "evidence_locator": "The patent record lists Vyoma Gajjar under Inventor for US11748453B2.", "checked_on": "2026-10-10" }, { "id": "ibm-research-adsa-2024", "claim": "Vyoma Rajeshkumar Gajjar is the named author of an invited talk published by IBM Research for ADSA 2024.", "source_url": "https://research.ibm.com/publications/navigating-the-complexities-of-generative-ais-ethical-social-and-legal-implications", "publisher": "IBM Research", "source_class": "publisher-record", "evidence_locator": "The IBM Research publication page identifies the work as an invited talk and names Vyoma Rajeshkumar Gajjar as author.", "checked_on": "2026-10-10" }, { "id": "ucla-rag-instructor", "claim": "Vyoma Gajjar is the listed instructor for UCLA Extension's Building Retrieval Augmented Generation course.", "source_url": "https://espa.unex.ucla.edu/computer-science/machine-learning-ai/course/building-retrieval-augmented-generation-rag-com-sci", "publisher": "UCLA Extension", "source_class": "publisher-record", "evidence_locator": "The public course schedule names Vyoma Gajjar in the Instructor field.", "checked_on": "2026-10-10" }, { "id": "servicenow-well-architected-review", "claim": "Vyoma Gajjar is an acknowledged contributor to ServiceNow's AI Agent Well-Architected Review.", "source_url": "https://www.servicenow.com/community/s/cgfwn76974/attachments/cgfwn76974/ceg-ai-coe-articles/59/2/ServiceNow_AI_Agents_Well-Architected_Review_v1.pdf", "publisher": "ServiceNow", "source_class": "publisher-record", "evidence_locator": "The acknowledgements on page 5 name Vyoma Gajjar as a contributor.", "checked_on": "2026-10-10" }, { "id": "ibm-mixture-of-experts-episode-51", "claim": "Vyoma Gajjar appeared as a guest on IBM's Mixture of Experts podcast episode 51.", "source_url": "https://www.ibm.com/think/podcasts/mixture-of-experts/o3-o4-mini-google-gemini-on-prem-nvidia-us-chip-manufacturing", "publisher": "IBM Think", "source_class": "publisher-record", "evidence_locator": "The episode page names Vyoma Gajjar in the guest list and throughout the transcript.", "checked_on": "2026-10-10" }, { "id": "ibm-techxchange-2024", "claim": "Vyoma Gajjar was listed as a presenter for an IBM TechXchange 2024 session on large language model development.", "source_url": "https://community.ibm.com/community/user/envirintel/blogs/mamatha-venkatesh/2024/10/17/meet-ibm-experts-at-ibm-techxchange-october-21-24", "publisher": "IBM Community", "source_class": "publisher-record", "evidence_locator": "The event post lists Vyoma Gajjar under session 2232, Putting Large Language Model development into the hands of AI developers and non developers.", "checked_on": "2026-10-10" } ] }