Patent 03 / Event replay ledger
01 / 11US10671515B1
↑↓  navigate  ·  all patents →
Siten Sanghvi / Granted June 2, 2020

The ledger that records.
Then replays.

A system for capturing every electronic event from a live source system onto a distributed ledger — and reconstructing those events, exactly as they occurred, in a test environment.

US10671515B1Patent number
2018-11-30Filed
1 yr 6 moTime to grant
19 Claims3 independent
21 CitationsForward citations
SCROLL TO EXPLORE THE SYSTEM
Visual patent explainer
01 / The problem

Production events are ephemeral.

Large distributed systems produce complex sequences of events. By the time a bug surfaces or a config change is considered, the exact sequence that triggered it is already gone.

The patent identifies three gaps that leave engineering teams unable to test against real history.

Three gaps in distributed system testing
01
Bugs vanish with the eventProduction event sequences are ephemeral. By the time a defect is noticed, the exact triggering sequence no longer exists to reproduce.
GONE
02
Synthetic tests miss real patternsScripted load and integration tests approximate real event distributions but cannot capture the complexity of actual production sequences.
APPROXIMATE
03
Config changes are a production gambleThere is no way to validate how a proposed configuration change would have handled the same historical events without changing the live system.
RISKY
02 / The inventive move

One ledger. Two modes.

The invention positions the distributed ledger as both recorder and replay engine. Record mode captures every event from the live source system. Playback mode reconstructs that exact sequence in a test environment.

The ledger is not just a log. It is the source of truth for both capturing history and replaying it.

Dual-mode ledger / record and playback
RECORD MODE
  • Monitor source system
  • Capture events with data types
  • Hash each data type separately
  • Write to distributed ledger
PLAYBACK MODE
  • Read from distributed ledger
  • Construct testing dataset
  • Simulate events in test env
  • Compare results to production
The same ledger serves both modes. Recording and playback are not separate systems — they share the chain as the source of truth.
03 / Architecture

From source to simulation.

A controller bridges the source system, the distributed ledger, and the testing environment. It monitors live events, records them onto the ledger, then reconstructs them for simulation.

The controller also manages selective cloning — choosing which section of the ledger to reconstruct — and stores software alongside the event data.

System architecture / four zones
01 / SOURCE SYSTEMProduction environment generating electronic events in real time.
02 / CONTROLLERMonitors, records, constructs testing dataset, and orchestrates simulation.
03 / DISTRIBUTED LEDGERImmutable, type-bucketed record of all events. The source of truth for both modes.
04 / TESTING ENVIRONMENTReceives the reconstructed event sequence and simulates outcomes.
04 / Typed blocks and hashing

Each data type gets its own hash.

When recording, the controller buckets event data by type within each block. It generates a separate hash for each data type, then stores those hashes individually in the modified block header.

This type-addressable structure is what makes selective reconstruction possible — each data type can be independently verified and replayed.

Modified block header / per-type hash storage
USER_AUTHH: 4ac1e8f2
TRANSACTIONH: 7bf29301
CONFIG_CHANGEH: 9d126fa4
SYSTEM_LOGH: b874c012
Block headers are modified to store separate hashes per data type — turning the ledger into a type-addressable event store, not just a sequential log.
05 / Selective clone and endblock value

Replay a window.
Not all of history.

The system can selectively clone a section of the distributed ledger. An "endblock value" designates the sequential blocks to be reconstructed in the testing environment.

This scopes replay to a specific time window — isolating the exact event sequence that led to a defect, a configuration change, or a specific operational scenario.

Selective clone / endblock demarcation
BLOCK 01EventsH:a7f1
BLOCK 02EventsH:b2c4
BLOCK 03STARTH:c891
BLOCK 04CloneH:d17f
BLOCK 05ENDBLOCKH:e9a3
TEST ENVRebuild03 → 05
source chaincloned sectiontarget environment
The endblock value is the chapter marker: reconstruct exactly blocks 03 through 05, leave the rest on the source chain.
06 / Configuration testing

Same events. Different configs.

A dependent claim covers replaying the same event sequence while applying different configuration changes. The system simulates the events first under Config A, then under Config B.

This turns the ledger into a regression harness for configuration — validate proposed changes against real production events before anything touches the live system.

Config regression harness / same events, two paths
REPLAYBlocks 03–05 reconstructed in test env
Same event sequence from production ledger
CONFIG A (CURRENT)

Simulate the production event sequence with the current configuration as baseline.

CONFIG B (PROPOSED)

Simulate the same event sequence with the proposed configuration change. Compare divergence.

Express: the patent claims configuration data changes during simulation. This pattern validates config before production deployment.
07 / Software capture

The ledger stores the software too.

A dependent claim describes storing the actual software associated with the source system onto the distributed ledger — alongside the events. During replay, the controller retrieves and implements that exact software version in the testing environment.

Events are replayed against the software that was actually running, not a current approximation of it.

Software-anchored replay / version-accurate simulation
LEDGER Block 04 contents
EVENT_DATA · HASH_PER_TYPE
SOFTWARE_ARTIFACT · v2.4.1
RETRIEVE Software artifact extracted from ledger
Exact binary version stored at record time
IMPLEMENT Software deployed in testing environment
Events processed using v2.4.1 — not the current version
SIMULATE Events replayed through version-accurate software
Defect reproduced in the exact context it occurred
08 / Practical applications

Where a replay ledger earns its keep.

The patent expressly claims production event recording and simulation in a testing environment. The same architecture suggests broader applications wherever real event sequences must be preserved and replayed faithfully.

Express = explicitly claimed. Inferred = reasonable extrapolation from the mechanism.

Application landscape
ExpressProduction bug replayCapture the exact event sequence that caused a defect and replay it in isolation.
ExpressConfig regression harnessReplay production events against a proposed config change before any live deployment.
InferredPre-deploy software validationStore a new software version on the ledger and replay real events to validate behavior before release.
InferredCompliance audit trailThe type-bucketed, hashed ledger provides an immutable, verifiable record of all system events.
InferredDisaster recovery dry runReplay a critical operational sequence in a test environment to validate recovery procedures.
InferredQA automation seedingSeed automated test suites with real event patterns captured from production ledger sections.
09 / Referenced by later work

Cited by IBM, Samsung, Visa International — and 9 more.

Google Patents lists 21 forward citations for this patent, spanning enterprise infrastructure, fintech, supply chain, and blockchain applications.

A forward citation shows relevance to later work. It does not imply commercial use or licensing.

Companies whose patents cite this work / checked June 10, 2026
Service virtualizationWELLS FARGO BANK, N.A. / US11372747B1Granted 2022
Ledger-based disaster recoveryINTERNATIONAL BUSINESS MACHINES / US11599431B2Granted 2023
Database optimized disaster recovery testingINTERNATIONAL BUSINESS MACHINES / US11003523B2Granted 2021
Data isolation in a blockchain networkADVANCED NEW TECHNOLOGIES CO. LTD / US11106817B2Granted 2021
Asynchronous updates in virtual distributed ledger networksJIO PLATFORMS LIMITED / US11762842B2Granted 2023
Distributed ledger based document image extracting and processingSPEEDCHAIN INC / US11616816B2Granted 2023
Integrated record of asset usage, maintenance, and conditionCATERPILLAR INC / US12093029B2Granted 2024
Data metrics compatibility in enterprise data warehouseCHIME FINANCIAL INC / US12099514B2Granted 2024
Immutable data structures for verifying financial dataINVENIAM CAPITAL PARTNERS INC / US12231566B2Granted 2025
Bot deterrence using cryptographyVISA INTERNATIONAL SERVICE ASSOCIATION / WO2025005916A1Published 2025
Partial ledger in blockchain networkSAMSUNG ELECTRONICS CO., LTD / US12621176B2Granted 2026
Bridging cross-platform network transfersCHIME FINANCIAL INC / US12572903B2Granted 2026
21 total forward citations as of June 10, 2026. Spanning enterprise, fintech, supply chain, AI, and blockchain infrastructure.
10 / Patent record

A quiet grant. A durable mechanism.

US10671515B1 codifies a general-purpose infrastructure pattern: use a distributed ledger not just as an audit log, but as the substrate for faithful event replay. Capture once, simulate many times.

Nov 2018
Priority Filed
US16/206,104 filed
priority date established
1 yr 6 mo
Jun 2020
Granted
USPTO granted US10671515B1
19 claims, 3 independent
~18 years
Nov 2038
Expires
Anticipated expiration
per Google Patents
Read the granted patent View the application publication Browse Siten Sanghvi patents
End / Patent 03