decagondev

AI Transparency & Trust Framework

Table of Contents

• Introduction
• Transparency Principles
• Trust-Building Beyond Transparency
• Error Management Framework
• Implementation Guide
• Communication Templates
• Measurement & Success Metrics

decagondev

Find the Unique Number in Triple Array

Problem Description

You are given an array of integers where every element appears exactly three times, except for one element which appears exactly once. Find that unique element.

Return that single element that appears only once.

Examples

decagondev

AI-Powered Apps

1. App: Startup Funding Deck Fixer

• Avatar: First-time founder
• Problem: Pitch deck too vague or unconvincing
• Solution: Upload your deck or paste text, get suggestions per slide to improve clarity, traction storytelling, and investor appeal.
• Crew AI Roles:
  • 🧠 Clarity Coach Agent — rewrites slides in plain, sharp English
• 📊 Investor Agent — checks for traction, TAM, GTM strategy and provides red flags

decagondev

Product Requirements Document: Startup Funding Deck Fixer

Overview

The Startup Funding Deck Fixer is an AI-powered tool designed to help first-time founders improve their pitch decks for investor presentations. Using specialized AI agents, the app analyzes pitch decks and provides actionable feedback to enhance clarity, storytelling, and overall investor appeal.

User Avatar

Primary User: First-time startup founder

• Limited experience with investor pitches
• May have technical expertise but struggles with investor communication
• Needs guidance on pitch deck best practices

decagondev

🚀 AI Security Learning Path Tracker

Fastest Path Summary (10-14 weeks full-time)

Phase	Topic	Time Estimate	Status	Notes
1	Networking + Linux + Cyber basics	2-3 weeks	Not Started
2	AI & ML fundamentals	1-2 weeks	Not Started
3	AI/ML threat landscape	2-3 weeks	Not Started
4	Deployment + LLM security	2-3 weeks	Not Started

decagondev

TOM TARPEY

Principal AI Engineer | Security & Database Analyst | Infrastructure & Software Architect | CS Consultant

+447513010774 | tomtarpeydev@gmail.com | LinkedIn: tom-tarpey-38594455 | GitHub: decagondev

TECHNICAL SKILLS

AI / ML / LLM:

• Large Language Models: GPT-4, Claude, LLaMA, Mistral, Falcon, Gemma, QLoRA, LoRA, PEFT
• LLM Fine-tuning: Custom dataset preparation, sentiment tuning, prompt engineering, hyperparameter optimization
• Quantization & Micro-LLMs: QLoRA/LoRA techniques for embedded & low-powered hardware

decagondev

Vehicle Valuation and Repair Cost Estimate - Chevrolet Orlando LT VCDi Auto (2011)

Overview

• Registration: AF61NHD
• Make/Model: Chevrolet Orlando LT VCDi Auto
• Fuel Type: Diesel
• Transmission: 6-Speed Auto
• First Registration: 24 October 2011
• Mileage (as of May 2025): 137,507 mi

decagondev

JavaScript: Object Properties, Value vs Reference, and Copying Objects

---

🔑 1. Object Properties

Objects in JavaScript are collections of key-value pairs. Each key is a property of the object.

const person = {

decagondev

PROBLEM SET

---

Problem 1: Two Sum Sorted (Array)

Difficulty: Easy
Category: Array

Description:

decagondev

RAG System Flow Diagram

This document illustrates the flow of a Retrieval-Augmented Generation (RAG) system implementation using LangChain.

System Overview

The RAG system combines document retrieval with language model generation to provide contextually relevant answers to user questions.

Flow Diagram