DCT
1:21-cv-01700
Valyant Ai Inc v. Hi Auto Inc
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: Valyant AI Inc (Colorado)
- Defendant: Hi Auto Inc (Delaware) and E La Carte Inc d/b/a Presto (Delaware)
- Plaintiff’s Counsel: Ashby & Geddes; Haynes & Boone LLP
- Case Identification: 1:21-cv-01700, D. Del., 02/09/2022
- Venue Allegations: Venue is asserted in the District of Delaware based on Defendants' incorporation in the state, their registration of agents for service of process, and their alleged business activities and acts of infringement within the district.
- Core Dispute: Plaintiff alleges that Defendants’ conversational artificial intelligence systems, used for automating drive-thru ordering, infringe a patent related to an AI-powered order processing system that uses human auditors for exception handling.
- Technical Context: The technology involves the application of artificial intelligence, speech recognition, and natural language processing to automate customer order-taking in the quick-service restaurant (QSR) industry.
- Key Procedural History: The complaint alleges that Plaintiff provided Defendant Hi Auto with actual notice of the patent-in-suit and its alleged infringement on June 15, 2021, and later provided a claim chart on October 13, 2021, following a request from Hi Auto's counsel. These pre-suit communications are asserted as a basis for willful infringement.
Case Timeline
| Date | Event |
|---|---|
| 2017-03-29 | '706 Patent Priority Date |
| 2019-01-01 | Plaintiff's technology first used in QSRs (alleged as "at least as early as 2019") |
| 2020-03-17 | '706 Patent Issue Date |
| 2021-04-01 | Defendant Hi Auto incorporated (approximate date) |
| 2021-06-15 | Plaintiff sent first notice letter to Hi Auto |
| 2021-07-21 | Plaintiff sent further correspondence to Hi Auto |
| 2021-10-13 | Plaintiff sent requested claim chart to Hi Auto |
| 2021-11-01 | Defendant Presto investor presentation citing Hi Auto data (approximate date) |
| 2022-01-01 | Defendant Presto announced major rollout with Checkers (approximate date) |
| 2022-02-09 | Complaint Filing Date |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 10,592,706 - "Artificially Intelligent Order Processing System"
- Patent Identification: U.S. Patent No. 10,592,706, "Artificially Intelligent Order Processing System," issued March 17, 2020.
The Invention Explained
- Problem Addressed: The patent's background section describes prior art speech recognition systems for retail as inefficient and unreliable, noting they often have limited vocabularies, struggle with accents or speech speed, and fail to recognize customer frustration, leading to missed sales opportunities and poor customer experience (’706 Patent, col. 2:4-29).
- The Patented Solution: The invention is an ordering system that uses artificial intelligence for speech-to-text conversion and natural language processing, but critically integrates a human "auditor" into the process (’706 Patent, Abstract; col. 6:36-44). This auditor, who can be located off-site, reviews the AI-processed order against the live customer audio feed to detect and correct errors, handle exceptions, or take over the transaction entirely if the AI fails or the customer becomes frustrated (’706 Patent, col. 8:21-43). The system can operate in a "training mode," an "auditor-assist mode," or an "autonomous mode" (’706 Patent, col. 6:30-34).
- Technical Importance: The invention's human-in-the-loop architecture for both training and real-time exception handling was designed to overcome the accuracy and reliability shortfalls of purely automated systems of the time (’706 Patent, col. 3:4-14).
Key Claims at a Glance
- The complaint asserts independent Claim 1 of the ’706 Patent (Compl. ¶¶ 14, 52).
- The essential elements of Claim 1, a method claim, are:
- providing an audio stream of a customer order
- providing an order processor with a speech recognition module trained with AI
- converting words in the audio stream to text
- processing the text to identify words based on previous training
- providing a natural language processor (NLP) with order assembly and exception detection capabilities to create or modify an order
- generating an order with the NLP
- alerting an off-site auditor of detected exceptions via a data connection
- deactivating the order processor with the auditor in response to exceptions
- prompting an on-site employee to engage the customer and complete the order
III. The Accused Instrumentality
Product Identification
- The complaint accuses Hi Auto's "conversational AI drive-thru technology" and Presto's "Presto Voice" software (Compl. ¶¶ 16, 30). The complaint alleges that Presto's product is the same as, or licenses technology from, Hi Auto (Compl. ¶¶ 37-38, 43).
Functionality and Market Context
- The accused products are described as AI-powered systems that automate the customer ordering process for QSR drive-thrus (Compl. ¶¶ 16, 31). The complaint alleges, based on Defendants' marketing materials, that the software can "process midsentence changes, handle multiple requests, understand accents, and work with 40 complex menus" (Compl. ¶22). A key functionality alleged is the use of an "offsite call center or a systematic back up" to manage detected exceptions or errors and permit human intervention (Compl. ¶25). The complaint includes a screenshot from a demonstration video showing the system displaying a digitally converted text order (Compl. ¶21). The complaint further alleges the technology has been deployed at QSRs, including Lee's Famous Recipe Chicken and Checkers (Compl. ¶¶ 28, 34).
IV. Analysis of Infringement Allegations
The complaint does not include the referenced Exhibit B claim chart (Compl. ¶54). The following table summarizes the infringement allegations for Claim 1 based on the narrative provided in the body of the complaint.
’706 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| providing an audio stream of an order of a customer positioned on a site; | The accused software receives a customer's order via an audio stream at a QSR drive-thru. A demonstration video is cited as showing this functionality. | ¶19 | col. 4:30-36 |
| providing an order processor having a speech recognition module trained using artificial intelligence programs; | The accused software is advertised as using "AI" or "artificial intelligence" and is said to learn to improve its order accuracy. | ¶21 | col. 4:3-6 |
| converting a word or words in the audio stream to text using the speech recognition module; | A demonstration video allegedly shows a customer order being "digitally" converted from audio to a text order by the accused software. The complaint provides a screenshot of this displayed text order. | ¶20, ¶21 | col. 4:37-43 |
| processing the text communication with the speech recognition module to identify a word or words in the text of the converted audio stream according to a previous spoken word training; | The accused software allegedly uses training to "understand accents" to identify words spoken by customers. | ¶22 | col. 4:44-48 |
| providing a natural language processor having order assembly capabilities and exception detection capabilities... | The software is alleged to use a natural language processor to create digital textual orders and to drive "exception detection by call center backups." | ¶23 | col. 4:49-54 |
| generating an order with the natural language processor; | The software is alleged to use a natural language processor to generate orders from customers' spoken words, as seen in demonstration videos. | ¶24 | col. 4:55-56 |
| alerting an auditor of detected exceptions in the order, the auditor located off-site and connected to the order processor via a data connection; | The software is allegedly configured to alert an "offsite call center or a systematic back up" to manage exceptions or errors. A demonstration video screenshot allegedly shows a "data connection box" installed in a QSR. | ¶25 | col. 4:57-61 |
| deactivating the order processor with the auditor in response to the detected exceptions in the order; and | It is alleged on information and belief that the software is configured to alert an offsite center, which deactivates the order processor in response to a detected exception. | ¶26 | col. 4:62-64 |
| prompting an on-site employee to engage the customer and complete the order. | The software allegedly provides digital orders to screens/monitors, which "prompts the on-site employees to fulfill the order." | ¶27 | col. 4:65-67 |
- Identified Points of Contention:
- Scope Questions: A primary question will be whether the Defendants' alleged "offsite call center or a systematic back up" (Compl. ¶25) meets the claim requirement of an "auditor." The resolution will depend on whether the functions of Defendants' backup system align with the functions described for the "auditor" in the patent's specification, such as real-time review and comparison of the AI-generated order to the customer's audio stream (’706 Patent, col. 8:21-43).
- Technical Questions: The complaint alleges "deactivating the order processor with the auditor" based on its investigation and "information and belief" (Compl. ¶26). A key factual question will be what evidence exists to show that an off-site human's response to an exception alert actively "deactivates" the AI processor, as opposed to simply correcting an order after the fact. The mechanism for "prompting an on-site employee" will also require factual development to determine if displaying an order on a screen constitutes the active step required by the claim.
V. Key Claim Terms for Construction
The Term: "auditor"
Context and Importance: This term is central to the patent's asserted point of novelty—the human-in-the-loop exception handling. The infringement case may depend on whether Defendants' "offsite call center" or "systematic back up" (Compl. ¶25) can be properly characterized as an "auditor." Practitioners may focus on this term because it is not explicitly defined with a simple dictionary definition, but is instead described by its function throughout the specification.
Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The patent does not limit the "auditor" to a single individual and describes the auditor as being "physically located off-site" (’706 Patent, col. 6:40-41), which could support an argument that a remote call center falls within its scope.
- Evidence for a Narrower Interpretation: The specification repeatedly describes the auditor performing specific, active functions: "review the order to detect errors or faults, such as by comparing the processed order in real time to the audio stream" (’706 Patent, col. 6:41-45), approving or editing an intent (col. 6:50-55), and having the ability to "take-over the order processing system" (col. 8:29-32). A defendant may argue that a simple backup system that only fields escalated problems does not meet this functional definition.
The Term: "deactivating the order processor with the auditor"
Context and Importance: This term describes a specific, interactive step between the auditor and the AI system. The allegation for this element is based on "information and belief" (Compl. ¶26), suggesting it will be a point of factual dispute. Proving this interaction occurred will be critical for the Plaintiff.
Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: "Deactivating" could be interpreted broadly to mean any action by the auditor that causes the AI to cease processing a specific order, such as taking control of the transaction, which the patent contemplates (’706 Patent, col. 8:30-32).
- Evidence for a Narrower Interpretation: The language may be construed more narrowly to require a specific command that formally switches the processor from an active to an inactive state. The specification mentions the auditor can "switch the artificially intelligent order processing system off to pass control to on-site employee(s)" (’706 Patent, col. 6:55-58), which suggests a more deliberate and complete disabling action.
VI. Other Allegations
- Indirect Infringement: The complaint alleges induced infringement under 35 U.S.C. § 271(b), asserting that Defendants knowingly instruct their QSR customers on how to implement and use the accused software in a manner that directly infringes Claim 1 of the ’706 Patent (Compl. ¶¶ 59, 61).
- Willful Infringement: The complaint pleads willfulness based on alleged pre-suit knowledge. It states that Plaintiff sent Hi Auto a letter identifying the ’706 Patent and the accused technology on June 15, 2021, and provided a claim chart on October 13, 2021, but alleges Defendants continued their infringing activities (Compl. ¶¶ 44-47, 55).
VII. Analyst’s Conclusion: Key Questions for the Case
- A core issue will be one of definitional scope: can the term "auditor," which the patent describes as performing specific real-time review and override functions, be construed to cover the Defendants' alleged "offsite call center or a systematic back up"? The outcome of claim construction for this term and its associated action steps will be pivotal.
- A key evidentiary question will be one of operational proof: what evidence will emerge in discovery to substantiate the complaint's "information and belief" allegations regarding the internal operations of the accused AI? Specifically, the case will likely turn on factual proof of how the system detects exceptions, what information is transmitted to the off-site center, and what precise mechanism constitutes "deactivating the order processor with the auditor."
- A third question relates to joint liability: what is the exact nature of the relationship between Hi Auto and Presto? The complaint alleges a partnership and licensing based on public statements, and the court will need to determine whether the alleged facts are sufficient to hold both parties liable for the same acts of infringement.