Season's Savings | ₹5/mo for 3 months

₹199/month after first 3 months. Renews automatically. Terms apply.
How to Build Your Own Private AI Assistant That Works Like an Employee cover art

How to Build Your Own Private AI Assistant That Works Like an Employee

How to Build Your Own Private AI Assistant That Works Like an Employee

Listen for free

View show details
About this listen
How can businesses use artificial intelligence without exposing confidential company data or paying expensive monthly AI subscription and token fees? One increasingly powerful solution is private AI: running your own AI models locally on your company's hardware. On this episode of Real AI Use Cases Business Owners Roundtable, hosts Elizabeth Gearhart and Richard Gearhart, Esq. talk with Matt Celia of LightSail VR about how he is using locally hosted AI systems to automate business operations, manage complex production schedules, protect sensitive information, and even simplify family life. Matt explains how he built a private AI assistant that acts like an employee, working around the clock to monitor Slack channels, track project schedules, organize stakeholder feedback, monitor computer security and storage, and automatically update production calendars when deadlines change. By running AI models on his own hardware, he can avoid ongoing token costs and keep sensitive client information within his own network. The conversation also explores how Matt created Dadbot, a personal AI assistant that gathers information from his children's multiple school and activity platforms and turns it into a printed weekly newspaper with schedules, homework reminders, activities, and even dad jokes. Other real-world AI applications discussed include using ChatGPT to research and purchase a car, building a private company knowledge base, creating an AI-powered second brain with Obsidian, and using local large language models (LLMs) to organize and connect information. You'll also hear an important discussion about data privacy, open-source AI, intellectual property, and why open source doesn't necessarily mean free. If you're a business owner wondering how to use AI to improve productivity, protect proprietary information, reduce costs, and build a competitive advantage, this episode offers practical ideas you can start exploring today. Guest: Matt Celia, LightSail VR — https://lightsailvr.com/ Hosts: Elizabeth Gearhart and Richard Gearhart, Esq. TL;DL – What You'll Learn How to build a private AI assistant that works like a 24/7 employee. Why businesses are running large language models (LLMs) locally instead of relying entirely on public AI platforms. How local AI can protect confidential company information and sensitive client data. How to eliminate ongoing AI token costs by running models on your own hardware. How AI can automate project management, production schedules, team communications, and stakeholder feedback. How to create a company knowledge base that remembers processes, expertise, and institutional knowledge. How Matt uses Dadbot to organize family schedules, school communications, homework, and activities. How ChatGPT can simplify complicated purchasing decisions, including researching and financing a new car. How Obsidian and AI can work together to create a searchable second brain. Why understanding data ownership, open-source software, and intellectual property matters as AI evolves. Frequently Asked Questions 1. What is private AI, and how does it work for businesses? Private AI allows businesses to run artificial intelligence models on their own computers or servers rather than relying entirely on public AI platforms. This gives companies greater control over their data, privacy, and how their AI systems are used. 2. Can a business run its own AI without sending confidential data to ChatGPT or other public AI platforms? Yes. Busi... Chapters (00:00:00) - Introduction to the Roundtable(00:00:23) - Immersive VR Production Challenges(00:02:06) - Post Production Bot and Local AI Processing(00:03:14) - Cost and Setup of Local LLM Hardware(00:04:56) - Private LLM vs Shared Files(00:05:20) - Dadbot: AI for Family Organization(00:07:40) - AI for Car Shopping and Financing(00:09:30) - Private LLMs for Business Knowledge Bases(00:10:14) - Networking Computers Without Internet via Tailscale(00:12:14) - Knowledge Databases and CEO Decision Modeling(00:12:53) - Karpathy's LLM Wiki Skill and Second Brain Concept(00:14:36) - Obsidian as a Second Brain Tool(00:15:15) - Data Ownership and Digital Literacy(00:15:58) - Open Source vs Proprietary AI Debate(00:17:45) - Understanding Open Source Licensing(00:18:34) - AEO, GEO, and SEO for AI Answer Engines(00:19:23) - Episode Wrap-up and Preview
adbl_web_anon_alc_button_suppression_t1
No reviews yet