Search
Close this search box.
Search
Close this search box.

HoundDog: Helping Developers Secure and Prevent Personally Identifiable Information from Leaking

Image Credits: iStock.com/JacobWackerhausen
By: Headliners News / May 30, 2024

Meet HoundDog.ai, a pioneering tech startup focused on preventing the leakage of personally identifiable information (PII) through developers’ code, has recently emerged from stealth mode. The company announced a successful $3.1 million seed funding round, led by E14, Mozilla Ventures, and ex/ante, alongside various angel investors. Distinct from other scanning tools, HoundDog.ai employs both traditional pattern matching and advanced large language models (LLMs) to scrutinize the actual code being written for potential vulnerabilities.

Founded by Amjad Afanah, who has a notable background with startups like DCHQ (acquired by Gridstore, later HyperGrid) and apisec.ai, HoundDog.ai was inspired by his experiences at data security startup Cyral. While at Cyral, the feedback from security and privacy teams was that current solutions were too reactive and couldn’t keep pace with changes in the codebase.

HoundDog.ai aims to address these concerns by integrating more closely with the development process. Positioned within the continuous integration workflow, the tool identifies potential data leaks before code merges, focusing on the code itself rather than the resulting data flow.

This proactive approach ensures that if developers start collecting sensitive information, like Social Security numbers, HoundDog.ai will alert both the development and security teams promptly, potentially averting significant issues.

Currently, the service supports a variety of languages and query types, including Java, C#, JavaScript, TypeScript, SQL, GraphQL, and OpenAPI/Swagger, with Python support on the horizon. The increasing reliance on AI-generated code makes tools like HoundDog.ai crucial. Ensuring the security of AI-generated code is essential, and HoundDog.ai is indispensable for securing PII early in the development cycles.

While HoundDog.ai leverages OpenAI’s models for its operations, users concerned about privacy can opt for traditional scanning methods to ensure their code remains within private repositories. This flexibility underscores the startup’s commitment to security.

Moreover, HoundDog.ai offers substantial compliance benefits by automating the generation of records of processing activities (RoPA) using generative AI. Importantly, only tokens identified by the scanner are sent to OpenAI, not the actual source code.

The company provides a limited free plan, with premium plans starting at $200 per month for scanning up to two repositories, offering a scalable solution for startups looking to enhance their code security and reduce compliance costs.

What do you think?

2 People voted this article. 2 Upvotes - 0 Downvotes.
Please Share This