Skip to content
View KaramelBytes's full-sized avatar

Highlights

  • Pro

Block or report KaramelBytes

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.

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
KaramelBytes/README.md

Twenty years in technology. MBA candidate at LSU Shreveport (accounting & finance).

The tech background doesn't go away — it just works on different problems now.


How I learn

I build things to understand them. Older repos cover AI-directed development and platform architecture; newer ones are interactive tools I build to internalize concepts before coursework starts — an LBO value-creation bridge, a dual-perspective accounting simulator, and similar.

The best way to understand a model is to build one.


Right now

  • 🎓 MBA candidate — LSU Shreveport, accounting & finance; studying for CFI FP&A
  • 🐍 Building graduate coursework learning apps in Python + Streamlit (accounting, finance, and strategic marketing analysis)
  • ⚡ Internal projects in TypeScript and Supabase
  • 🔌 Developing Model Context Protocol (MCP) servers and clients to the Anthropic spec
  • 🏠 Self-hosted LibreChat deployment with custom agents, MCP tools, on-server RAG using local embedded models, and tuned token efficiency
  • 🌐 LLM routing through OpenRouter with a zero-data-retention (ZDR) workflow, rotating frontier and preview models for task-fit and cost efficiency, plus Hugging Face for local inference

Tech stack

Building with: Python, TypeScript, Streamlit

Data & backend: Supabase

Running on: Vercel, Cloudflare, self-hosted Debian (Docker)

Automation & integration: n8n, MCP

Working with LLMs via: OpenRouter (primary routing, ZDR), Hugging Face (local inference), and direct APIs — rotating frontier models such as Hy3, GLM 5.1, and Kimi K2.6, plus DeepSeek v4 and others


Selected projects


🔗 jeremiahhenning.com · LinkedIn

Pinned Loading

  1. accounting-learning-sim accounting-learning-sim Public

    An interactive Streamlit simulator teaching business scenarios from dual perspectives: the Auditor's View (GAAP) and the CFO's View (Managerial Accounting).

    Python

  2. LBO-learning-sim LBO-learning-sim Public

    An interactive Streamlit dashboard that decomposes Leveraged Buyout (LBO) equity value creation into core financial and operational drivers.

    Python