AHA : Scalable Alternative History Analysis for Operational Timeseries Applications Conference: Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1; August 9–13, 2026; Jeju Island, Republic of Korea. Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Disc
Harshavardhan Kamarthi, Harshil Shah, Henry Milner, Sayan Sinha, Yan Li, B. Aditya Prakash, Vyas Sekar · 2026
Read AHA : Scalable Alternative History Analysis for Operational Timeseries Applications Conference: Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1; August 9–13, 2026; Jeju Island, Republic of Korea. Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Disc by Harshavardhan Kamarthi, Harshil Shah, Henry Milner, Sayan Sinha, Yan Li, B. Aditya Prakash, Vyas Sekar (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-04432.txt.
Description
Harshavardhan Kamarthi and six co-authors build a time machine for operational data in a 2026 arXiv paper. Companies that stream video or run networks collect rich per-session quality metrics, and analysts constantly want to replay history differently: test a new anomaly detector, retune an alert, score a new algorithm. Warehouses, samples, and sketches either cost too much for that or cannot replay faithfully.
Their system, AHA, rests on three observations about such workloads. The statistics involved decompose neatly, only a few subpopulations are active among the many attribute combinations, and modern analytics databases aggregate in exploitable ways. On real datasets and production pipelines at a large video analytics company. AHA replayed with full accuracy across downstream tasks at up to eighty-five times lower total ownership cost than conventional methods, counting both computing and storage.
Catalog record
| CREATOR | Harshavardhan Kamarthi, Harshil Shah, Henry Milner, Sayan Sinha, Yan Li, B. Aditya Prakash, Vyas Sekar |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-04432