Michael Stattelman@michael_stattelman238

Microsoft Recall Open Sourced version. This project takes timed screenshots of your screen and adds metadata to each new image is saved and described using a PHI3 Vision model; this allows for searchable device history.

Purpose: The primary objective of employing Cogito is to enhance a gaming platform's ability to maintain a healthy and respectful community environment. By scanning chat logs and real-time conversations through an API, the system aims to identify and address potential instances of player terms violations. The system has the capability to actively monitors both static chat logs and live-streamed chats to ensure comprehensive coverage of user interactions. Static chat logs are historical records of conversations, while live-streamed chats represent ongoing interactions among players during gameplay or other activities. Linguistic Cues Detection: Cogito utilizes advanced natural language processing techniques through the Mixtral LLM to identify linguistic cues indicative of potential violations of player terms.

TEMPESTAS, powered by Falcons.AI's "Phi-2" model, represents the forefront of chaos engineering. Discover its proactive approach in identifying and mitigating potential system failures, ensuring unparalleled reliability for your applications. With the introduction of the H.A.N.D.S methodology (Hardware, Application, Network, Data, Security), TEMPESTAS revolutionizes chaos engineering by comprehensively stress-testing systems. Experience how TEMPESTAS, using H.A.N.D.S., identifies vulnerabilities, strengthening your infrastructure against unforeseen failures. This holistic approach provides a robust framework for enhancing system resilience against diverse challenges.

This new metric "Polypharmic risk Score " is a metric I am currently developing in an effort to be combined with the polygenic risk scores to provide a truly "Precision Healthcare" analysis. The polypharmic risk is that of over the counter medications taken over time and their potential tot introduce increased risks or adverse health implications. The solution is currently limited by the healthcare data hurdles I encountered and can be alleviated with proper access to such when provided to the model for analysis. To display the analystical use case for the FalconLLM model I used the AI71 API key and built that into the streamlit application. the application is currently limited to a set of common over the counter medications as well as common illnesses. This can be expanded with comprehensive medication related data as well as patient specific data to enhance the overall accuracy.