We envision IKI.AI as a next generation knowledge interface, blurring the boundaries between knowledge management and a proactive expert assistant

We believe that modern LLMs, RAG and Agentic algorithms combined with the professional community expertise would provide far more efficient ways of organizing and adopting professional knowledge.

We co-founded IKI because we faced information overload ourselves while browsing professional content and working with our teams.

The purpose of IKI.AI is to help professionals with knowledge storage, sharing, collaboration and peer-to-peer learning to facilitate new knowledge creation, ideas adoption and implementation through a proactive co-pilot.

There is a broad spectrum of applications and target audiences – professionals, students and teams.

The use cases are different, but the core principles are the same. Self-organizing, seamlessly searchable space to store your data and derive insights.

Book demo

Go to market strategies

Wearable

A number of stories and analysis of go to market strategies from the most successful startups like Slack, Udemy, Facebook - enjoy reading and picking the things that would work for you.

14 items

Ivan Ilin

Top 10 Design systems in 2024

Wearable

A number of stories and analysis of go to market strategies from the most successful startups like Slack, Udemy, Facebook - enjoy reading and picking the things that would work for you.

14 items

Max Orlov

Agents

Wearable

AI summary: 40+ publications and repositories with the most capable Agentic architectures & implementations.

27 items

Ivan Ilin

Advanced RAG

Wearable

AI summary: The collection primarily discusses advanced techniques and methodologies for enhancing research and information retrieval using AI models - retrieval augmanted generation or RAG. It covers key approaches to retrieving relevant context, ranking and resoning upon it with an LLM,plus introduces some frameworks for quality control and metrics.

41 items

Ivan Ilin

Publications

Published in Towards AI

Through Knowledge Sharing to Singularity, Accelerated By LLMs

AI summary: LLMs are one of the pinnacles of human knowledge that has transformative potential comparable to the Internet. How did we come to that? Why do knowledge sharing and knowledge flow play a crucial role in our world’s acceleration, and why are LLMs the lapis philosophorum that could bring us to the singularity?

Published in Towards AI

Through Knowledge Sharing to Singularity, Accelerated By LLMs

AI summary: LLMs are one of the pinnacles of human knowledge that has transformative potential comparable to the Internet. How did we come to that? Why do knowledge sharing and knowledge flow play a crucial role in our world’s acceleration, and why are LLMs the lapis philosophorum that could bring us to the singularity?

Published in Towards AI

Advanced RAG Techniques: an Illustrated Overview

AI summary: A comprehensive study of the advanced retrieval augmented generation techniques and algorithms, systemising various approaches. The article comes with a collection of links in my knowledge base referencing various implementations and studies mentioned.

Published in Towards AI

Advanced RAG Techniques: an Illustrated Overview

AI summary: A comprehensive study of the advanced retrieval augmented generation techniques and algorithms, systemising various approaches. The article comes with a collection of links in my knowledge base referencing various implementations and studies mentioned.

Published in Towards Data Science

Building a news aggregator from scratch: news filtering, classification, grouping in threads and ranking

AI summary: The idea behind this post is to show a reasonably simple approach one can implement in a couple of weeks to solve a real-world problem of creating a news aggregator like Google news or Yandex news showing the top news threads out of millions of news scraped all over the web.

Published in Towards Data Science

Building a news aggregator from scratch: news filtering, classification, grouping in threads and ranking

AI summary: The idea behind this post is to show a reasonably simple approach one can implement in a couple of weeks to solve a real-world problem of creating a news aggregator like Google news or Yandex news showing the top news threads out of millions of news scraped all over the web.

Digital Library

Where professional knowledge lives

"A digital library of the future - searchable, self-organising, with an AI assistant instantly fetching ideas, relations, helping with reasoning and ideation."