Hi, my name is

Micah Katz

AI Software Developer II @ SelectQuote

Projects

Workflow Manager UI

Created a full-stack UI for managing SelectQuote AI call workflows.

Next.js

React.js

Node.js

Agent Assist Desktop App

Developed an Electron app that records desktop audio, transcribes it, and extracts sentiment data to be displayed over time.

Electron

React.js

Whisper API

Custom ML Model

Shadcn

SelectQuote.ai

Created a full-stack UI for all SelectQuote employees to interact with multiple LLMs.

Next.js

React.js

Node.js

OpenAI API

Machine Learning Hardhat Alerts UI

Developed a live MVIE demo that was shown to notable clients and at worldwide conferences (MWC).

React

Node

Linux

MQTT

MVI Edge

Internal Retrieval Augmented Generation

Implemented Retrieval Augmented Generation on an internal UI. This allowed SelectQuote employees to ask questions about large knowledge bases.

Python

Langchain

DALLE API

Interactive Large Action Model

Implemented cutting-edge AI concepts to allow an LLM to choose between different custom actions intelligently. This allowed SelectQuote employees to query live internal data and static files.

Python

Langchain

DALLE API

Entity Extraction Neural Network

Created a neural network to reduce latency and cost in a production application.

Python

Tensorflow

Pandas

Sentiment Classification Neural Network

Created a neural network to reduce latency and cost in a production application.

Python

Tensorflow

Pandas

Team Solutions Dashboard

Developed a dashboard for my IBM team to do CRUD on our internal website containing markdown articles.

React

Node

Linux

API

CSS

5G System Status UI

Developed a draggable flowchart UI showing internal IBM 5G systems and their JSON data as tables.

React

Node

Linux

API

CSS

Experience

SelectQuote
IBM
JamFeed
6 Points Sci-Tech Academy
Anubis LLC
AI Software Developer II @ SelectQuote

Aug 2023 - Present

  • Led a team of developers to build a feature-rich editor for managing LLM phone call workflows.
  • Created neural networks that improved speed for a real-time production application 200-fold.
  • Implemented a Large Action Model that lets employees interact with a company database via LLM.
  • Integrated RAG, LLMs, and a LAM into a chat website for all SelectQuote employees.
Created with React.js, Next.js, TypeScript, and TailwindCSS