{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:WLOLTAU2ZEW4SFJGZ3Q4LZCXWN","short_pith_number":"pith:WLOLTAU2","schema_version":"1.0","canonical_sha256":"b2dcb9829ac92dc91526cee1c5e457b36cd1fa1db0a0b1044065bde22a754bf2","source":{"kind":"arxiv","id":"2407.12003","version":2},"attestation_state":"computed","paper":{"title":"Evaluation and Continual Improvement for an Enterprise AI Assistant","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.HC","authors_text":"Akash V. Maharaj, Horia Galatanu, Ken Russell, Kun Qian, Manas Garg, Nishant Kapoor, Rachel Hanessian, Sally Fang, Shivakumar Vaithyanathan, Uttaran Bhattacharya, Yunyao Li","submitted_at":"2024-06-15T06:00:23Z","abstract_excerpt":"The development of conversational AI assistants is an iterative process with multiple components. As such, the evaluation and continual improvement of these assistants is a complex and multifaceted problem. This paper introduces the challenges in evaluating and improving a generative AI assistant for enterprises, which is under active development, and how we address these challenges. We also share preliminary results and discuss lessons learned."},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2407.12003","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.HC","submitted_at":"2024-06-15T06:00:23Z","cross_cats_sorted":[],"title_canon_sha256":"cbd6981c05c911ce648cfaef52393710443d95592439e2c6f55a0de93f1bc9e5","abstract_canon_sha256":"cc6d8b9aafe4dbd73952d645ae0df4d885fc657b35eda36970b64742298ba6b6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:45:51.662981Z","signature_b64":"oLHghnlDLRvJiuponSKhYhGpo7Fx0Dt95W3gKFmFd38GaCt68dD4lOskL1vk0Plv7GWl+c10zOEc5fwI75IRAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b2dcb9829ac92dc91526cee1c5e457b36cd1fa1db0a0b1044065bde22a754bf2","last_reissued_at":"2026-07-05T09:45:51.662525Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:45:51.662525Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evaluation and Continual Improvement for an Enterprise AI Assistant","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.HC","authors_text":"Akash V. Maharaj, Horia Galatanu, Ken Russell, Kun Qian, Manas Garg, Nishant Kapoor, Rachel Hanessian, Sally Fang, Shivakumar Vaithyanathan, Uttaran Bhattacharya, Yunyao Li","submitted_at":"2024-06-15T06:00:23Z","abstract_excerpt":"The development of conversational AI assistants is an iterative process with multiple components. As such, the evaluation and continual improvement of these assistants is a complex and multifaceted problem. This paper introduces the challenges in evaluating and improving a generative AI assistant for enterprises, which is under active development, and how we address these challenges. We also share preliminary results and discuss lessons learned."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.12003","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2407.12003/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2407.12003","created_at":"2026-07-05T09:45:51.662585+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.12003v2","created_at":"2026-07-05T09:45:51.662585+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.12003","created_at":"2026-07-05T09:45:51.662585+00:00"},{"alias_kind":"pith_short_12","alias_value":"WLOLTAU2ZEW4","created_at":"2026-07-05T09:45:51.662585+00:00"},{"alias_kind":"pith_short_16","alias_value":"WLOLTAU2ZEW4SFJG","created_at":"2026-07-05T09:45:51.662585+00:00"},{"alias_kind":"pith_short_8","alias_value":"WLOLTAU2","created_at":"2026-07-05T09:45:51.662585+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN","json":"https://pith.science/pith/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN.json","graph_json":"https://pith.science/api/pith-number/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN/graph.json","events_json":"https://pith.science/api/pith-number/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN/events.json","paper":"https://pith.science/paper/WLOLTAU2"},"agent_actions":{"view_html":"https://pith.science/pith/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN","download_json":"https://pith.science/pith/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN.json","view_paper":"https://pith.science/paper/WLOLTAU2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.12003&json=true","fetch_graph":"https://pith.science/api/pith-number/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN/graph.json","fetch_events":"https://pith.science/api/pith-number/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN/action/storage_attestation","attest_author":"https://pith.science/pith/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN/action/author_attestation","sign_citation":"https://pith.science/pith/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN/action/citation_signature","submit_replication":"https://pith.science/pith/WLOLTAU2ZEW4SFJGZ3Q4LZCXWN/action/replication_record"}},"created_at":"2026-07-05T09:45:51.662585+00:00","updated_at":"2026-07-05T09:45:51.662585+00:00"}