2024 upped the game: AI in Effectus PreSale App

Reading time: 5 minutes

Contents

What is PreSale?

PreSale is a tool designed to enhance and optimize the sales process. It enables potential clients to complete a set of questions to streamline the selling pipeline and offers the option to schedule a meeting. Its purpose is to maximize efficiency, identify opportunities, and help sales teams make more informed decisions.

By leveraging the power of artificial intelligence (AI), PreSale allows salespeople to accurately identify potential customers, personalize interactions, and prioritize their efforts more effectively. In a world where data is abundant but attention is scarce, PreSale uses AI to analyze and process large volumes of information in real-time, extracting valuable insights that might otherwise go unnoticed. This leads to less time spent on repetitive tasks and a greater focus on activities that truly impact customer conversion.

One of the standout features of PreSale is its use of AI to suggest responses during the questionnaire. As clients fill out the form, our model provides helpful suggestions, ensuring that responses are as complete and accurate as possible. This not only speeds up the process but also helps in gathering high-quality data.

After gathering responses, PreSale uses AI to categorize leads based on their maturity level, interest, and potential fit. This categorization helps sales teams prioritize their efforts on the most promising leads, ensuring that their time and resources are spent where they can have the most impact.

To make the most of AI, it’s crucial to use effective prompts. Prompts are commands or queries given to AI models to generate relevant and contextual responses, and the way they are crafted can greatly influence the quality of the results. Creating clear, specific, and well-structured prompts is essential for achieving optimal outcomes.

To create more sophisticated and dynamic prompts, PreSale leverages a powerful tool called LangChain…

Before getting into how Langchain works, it is necessary to understand some basic concepts and terminology used in this context:

Chain: The fundamental unit of work in LangChain. It’s a sequence of steps that process prompts and generate responses.
Prompt: The input text that triggers the chain’s response.
LLM: Large Language Model, like GPT-3 or LLaMA, which is often used as a component within chains.
Tools: Additional components that can be integrated into chains, such as search engines, calculators, or databases.

LLMs

LLMs are the backbone of LangChain. They generate text, translate languages, write different kinds of creative content, and answer your questions in an informative way. There are many LLM providers, but amongst the most common ones are OpenAI (GPT-3, GPT-4), Hugging Face Transformers, and Google AI.

Prompts

Prompts act as instructions for the LLM. Well-crafted prompts can significantly influence the quality and relevance of the generated response. There are various techniques that can be implemented to generate a good, well-crafted prompt. Below we list some of the most common ones:

Clarity and Specificity: How to provide clear and concise instructions.
Contextual Information: Incorporating relevant background information.
Prompt Templates: Using pre-defined templates for common tasks.
Few-Shot Learning: Providing examples to guide the LLM’s understanding.

Basic code snippet to use Langchain

Install

On your terminal run the command:
yarn install langchain

Setup

2024 upped the game: AI in Effectus PreSale App 1

Basic reflection pattern

“The reflector is prompted to role play as a teacher and offer constructive criticism for the initial response. The loop proceeds a fixed number of times, and the final generated output is returned.” Open AI.

2024 upped the game: AI in Effectus PreSale App 2

Reflection actor

“Within reflection, the actor agent explicitly critiques each response and grounds its criticism in external data. It is forced to generate citations and explicitly enumerate superfluous and missing aspects of the generated response. This makes the content of the reflections more constructive and better steers the generator in responding to the feedback.” Open AI.