Reach out to me to seamlessly integrate these AI systems into your operations and stay ahead in technology.

Section 0: Introduction

Here, I will briefly elaborate on my plans in simple terms.

0.1 My Background

When I entered my Bachelor program in Computer Science, I realized we are mostly memorizing facts which shattered my picture of University-level education. So, I embarked on a journey of self-studying about Phylosophy, Cognitive Science, and Epistemology, mainly under the supervison of Dr. Milad Nouri who was a Professor at another University. I aimed at mapping the human mind to automate "Learning, Discovery, and Problem Solving" so that afterward, we all go on vacation and enjoy our lifes.

In "Learning" part, I realized how insifficient our understanding of our congnitive systems are. In "Discovery" I examined the source of our knowledge and gained a realistic picture of its limitations and how we may be able to push the boundaries. Most importantly, in "Problem Solving" I learnt about two theoretical and Practical ceilings (Godel Incompleteness Theorem & Incomputable Problems in Computer Science).

Long story short, after a meandering road, after knowking limitations and capabilities, I pursued my desire of automating mundane tasks by working on Languag Models and Chatbots.

0.2 Objectives

I am documenting the theory of Machine Learning and Chatbot systems on my other tutorials and pages. Here, I will focus on implementing and deploying these systems to bridge gaps in the market and deliver economic gains to stakeholders.

Section 1: Rule-based Chatbots

I have implemented a client-side rule-based chatbot with UI which is available on my repository^[1]. It's true that rule-based chatbots are simple but don't get me wrong. They resolve 70 percent of customer tickets, based on a recent statistic.

Section 2: Retrieval Augmented Generation (RAG) | Inventures Case Study

I implemented a RAG system and did not sweep anything under the rug so that we can have a realistic expectation of its capabilities. I am grateful for Javier whose repository^[2] tuaght me how to run localtunnel on Colab. He also answred my doubt on Graph-DBMS. The thumbnail photo is made by @SombilonStudios. I would like to emphesis that the only source of infromation for Inventures session is alberta Innovates website and if you want to use their data, you should check their most recent privary page^[3] .

2.1 Retrieval Augmented Generation RAG. Honest Review. No Cherry Picking.

2.2 RAG with Knowledge Graph, Neo4j, VectorDB, GPT Large Language Model, and RAPTOR

2.3 Run RAG LOCALLY | LM studio | nomic-ai embeddings | Mistral LLM | Neo4j Desktop macOS

2.4 HippoRAG Neurobiologically Inspired LongTerm Memory for Large Language Models | KG | PageRank

HippoRAG is a novel retrieval framework for large language models (LLMs) inspired by the hippocampal indexing theory of human memory. Developed by researchers from Ohio State University and Stanford University, it enhances the ability of LLMs to dynamically integrate and retrieve knowledge more effectively.

In my video, I start by introducing the content and outlining the main goals. I then explain the framework used to review the paper, followed by a discussion of the specific problem it addresses. I delve into the motivation behind the research and present the proposed solution. Next, I evaluate the results, reflect on my learning, and explore future research directions. I also provide a practical demonstration of running the code on Colab, and conclude with a summary of the key takeaways.

Section 3: Intent-detection & Slot-filling Chatbot

Coming soon!

Section 4: AI Tutor, A Wish Which Didn't Happen for My Generation.

Coming soon!

Section 5: Personal Mentor

Coming soon!

Your favorite usecase isn't among these sections? Reach out to me to make that happen.

References

1. https://github.com/Reza-Ardestani/ConversationalAI
2. https://github.com/javier-jaime/Tool-Crib/blob/master/LangChain/Information_retrieval_with_tools.ipynb
3. https://inventurescanada.com/privacy-policy/