Every AI application, interactive engine, and creative system featured on this site is an original invention designed, conceptualized, and built from scratch by Christopher Terrell. No templates, drag-and-drop software builders, or pre-packaged AI wrappers were ever used. Every application was uniquely hand coded using pure Python to transform complex data and narrative challenges into custom software solutions.

AI Accuracy Demonstration Model Overview: The greatest risk in adopting generative AI for business is unpredictability. Standard AI models rely on probability—they guess the next word, which often leads to costly, untrustworthy "hallucinations." This interactive demonstration proves that AI can be engineered to deliver 100% accurate, institutionally trusted results every single time. By executing a live, side-by-side comparison using a wide, unrestricted scope of global topics, it shows that even when a user enters a completely un-engineered, casual prompt, behind-the-scenes orchestration ensures the output matches rigid regulatory standards. Latest enhancements: The model was greatly enhanced to allow queries on all topics rather than a few narrow business ones. How it Works: The Unanchored Node: Processes the raw user prompt through a standard generative model layer without localized contextual constraints, demonstrating the high-variance baseline. The Orchestrated Node: Captures the same raw prompt and routing payload through a custom, behind-the-scenes middleware curation matrix. This layer dynamically enforces a strict, multi-stage Chain-of-Thought (CoT) reasoning path, forcing the model to explicitly execute mathematical boundary checks, temporal mapping, and logical verification vectors before compiling the final text response. Engineering & Creation: Every line of this framework was hard-coded from scratch using pure Python and deployed as an integrated web container. • State Management: Built leveraging Streamlit's reactive application framework for low-latency, real-time UI state mutations. • API Ingestion Nodes: Utilizes highly optimized OpenAI API communication protocols, passing custom programmatic system instructions to enforce deterministic parameter controls (e.g., locking temperature values at an absolute 0.0 vector to eliminate stochastic token selection variation). • Data Scrape Integration: Incorporates live REST API handshakes and direct Google search indexing to cross-reference historical temporal facts before generating the final compliance output. 



Click to try it

Note: Because these models are currently for demo purposes only, they may need to be "woken up" before loading.

The Copyright Compliance Engine Overview: Navigating music licensing is a legal minefield for creators. Sampling a hook or utilizing archival lyrics can lead to costly copyright lawsuits if the track’s legal status is miscalculated. This application functions as a reliable legal logic calculator engineered to automate the complex discovery phase of copyright clearance. By analyzing simple metadata or lyric snippets, the engine instantly evaluates the asset against strict United States statutory copyright thresholds. This tool demonstrates how custom-engineered AI can actively protect artists' rights—ensuring musicians do not unintentionally infringe on active copyrights, while giving them the confidence to freely utilize works that have safely entered the public domain. Latest enhancement: A lyrics search tab was added as well as links to the copyright holders How it Works: The system ingests data through three distinct tracks (Lyric Snippets, Title/Artist strings, or Quick Custom Entry). • Multi-Modal Data Anchoring: When a lyric phrase is entered, the engine routes the text through the Genius API to extract the master track parameters. If the primary database contains missing or unindexed timeline metadata, the engine instantly deploys a secondary background request via the Google Live Index Search gateway. • Deterministic Evaluation: Once the original publication year is isolated, a regex matrix extracts the temporal data vector and passes it to an isolated, zero-temperature OpenAI API node. The model evaluates the asset against an unyielding legal boundary condition: any work published on or before December 31, 1930, has surpassed its 95-year statutory term limit and resides unconditionally within the US Public Domain. Any work published on or after January 1, 1931, triggers an active restriction warning.

Engineering, Creation & Future Enterprise Vision: The entire architecture was hard-coded by hand using Python and deployed as a reactive web application container via Streamlit. The backend utilizes custom requests payload pipelines to seamlessly bridge crowdsourced music repositories with live Google search snippets. If an asset is flagged as copyrighted, the system dynamically generates authenticated industry destination paths—automatically providing the user with direct access to HFA Songfile for mechanical licensing and the unified ASCAP/BMI Songview Repository for synchronization rights research. While this current build serves as a robust functional foundation, the framework has been explicitly engineered to scale into a highly marketable, comprehensive enterprise B2B SaaS platform designed to completely automate intellectual property discovery for advertising agencies, film studios, and record labels. The roadmap includes: • Transitioning from Static Math to Dynamic RAG: Upgrading the current date-threshold logic to a dynamic Retrieval-Augmented Generation (RAG) pipeline, allowing the AI to actively read, parse, and interpret complex legal court precedents, fair-use defense records, and corporate catalog acquisitions. • Direct Rights-Directory Integration: Moving beyond public web indexing to establish secure database handshakes with the US Copyright Office, ASCAP, and BMI to track unbroken chains of title in real time. • Expanding to Multi-Modal Audio Parsing: Integrating acoustic fingerprinting models alongside text processing, allowing users to upload raw audio files to verify master sound recording provenance in tandem with written lyrics and prose. • Automated License Liquidation Flow: Advancing the output from descriptive compliance alerts to an actionable deal-flow utility that identifies the exact corporate rights-holders, calculates complex percentage splits, and generates the explicit routing paperwork required to secure mechanical or sync licenses instantly.

Click to Try It

Note: Because these models are currently for demo purposes only, they may need to be "woken up" before loading.