Ask Away AI

Product Design Bylaw 14.3 ... must have napkin sketch
user flow
Conceptual wireframe
Anatomy of an Insight Card

The Big Idea:

Every company has competitors. What do you want to know about yours?

The Problem:

Everyone loves the idea of Competitive Intelligence. Few enjoy the time it takes. Even fewer know how to get the essentials right.

it's always an awkward cold start (port over from where? start fresh? fresh with what?)
it's time consuming, often off the side of a PMM's desk
it's fraught with low-value, repeatable tasks
it's stored in docs and decks to decay in the dark recesses of a wiki
it's often reduced to ... "Hey, I need something. Can you send my prospect that feature comparison for me?"
which then leads to ... "Hey, this is old. Where is the updated one?"

The Solution:

Overview

At a high-level, users can learn about companies and products in a concise UI that is optimized to compare, contrast and extrapolate information. An obvious job for GPT's, but the special sauce is fine-tuning with unique "compete data" and tracking source data for changes.

UX

The premise of the UX starts with a prompt for any competitor company or product. The prompt returns a cascading set of Insight Cards ("Klues"); these contain distinct meaningful nuggets of truth broken out from the users prompt.

Agency is how Insight Cards act on the users behalf. If sources change, for example, the Insight is updated. Card content is as liquid as the information and data that derived it. In short, Agency takes care of a long-tail problem from Gen AI. Someone needs to maintain it, at scale, every month, week, day, hour ... minute. Continuous fresh, valid, trusted Insights that help sellers win deals is the Holy Grail we're aiming for and AI is our Indiana Jones.

Insight Cards: Comprised of three properties

a unique prompt (generated the Insight)
the sources used to generate the Insight & prompt
the Insight (content, core value prop), editable by humans

Insight Card Agency: A mechanism for automating when information has changed (e.g. the sources)

follow the chain of prompts leading to an Insight (value add)
as sources change, see how the insight changed
before and after, etc
automated freshness

Bonus Thinking:

As a team coming from Klue, the competitive intelligence lens is top of mind, but this approach would work across a myriad of topics, concepts, industries and verticals.

anything you want to effortlessly keep tabs on
knowledge bases that are disparate (redundancy, verbosity, misinformation)
use cases where an open-source LLM needs to be hyper-focused for a specific outcome
end-users just looking for a quick "good enough" answer

The Team

Comprised of the wonderful engineering talents of Julian and David, with support from Jesse, the team hails from Vancouver, Canada. Employed together at Klue Labs Inc., they are excited to bring back heads and hearts full of knowledge and inspiration for their teammates.

Disclaimer: Craft is a investor in Klue Double-Disclaimer: ChatGPT was not used in the creating of this submission. All drivel, rambling and long-windedness is proudly human.