Skip to content
View ashwinvis98's full-sized avatar
4 followers · 7 following

Block or report ashwinvis98

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ashwinvis98/README.md

Ashwin Viswamithiran

I am a security engineer in Denver. I build the platforms, data pipelines and automation behind a threat research team, so that what it learns about adversaries reaches our defenses without anyone carrying it over by hand. I also look for where AI can make that work faster and more dependable.

ashwinviswamithiran.com · Writing · LinkedIn · Résumé

Projects

  • opencti-ai-enrichment: an OpenCTI enrichment connector where a language model proposes entities and relationships, and deterministic code decides what reaches the knowledge graph.
  • adversarial-ai-cti: prompt injection and jailbreaks as vendor-neutral STIX 2.1 threat intelligence, so any STIX platform can ingest them.
  • promptlsh: a portable similarity digest for prompt attacks, so reworded jailbreaks correlate instead of reading as unrelated items.

Each project has a decision record on the site: what it does, what I chose and why, and where it falls short.

Upstream

  • OpenCTI connectors: support for every detection-rule pattern type and multi-rule documents in the Recorded Future connector.

Tools I work with

OpenCTI · STIX 2.1 · MITRE ATT&CK and ATLAS · Cortex XSIAM · IBM QRadar · Python · GraphQL · AWS and Amazon Bedrock

Pinned Loading

  1. adversarial-ai-cti adversarial-ai-cti Public

    Represent adversarial AI activity (prompt injection, jailbreaks) as STIX 2.1 threat intelligence.

    Python 1

  2. promptlsh promptlsh Public

    A portable similarity digest for prompt attacks, so reworded jailbreaks correlate instead of reading as unrelated items.

    Python 1

  3. opencti-ai-enrichment opencti-ai-enrichment Public

    An OpenCTI enrichment connector where the language model proposes entities and relationships and deterministic code decides what reaches the knowledge graph.

    Python