Today, let’s discuss the signs indicating it’s time to integrate AI into a product and whether artificial intelligence can deliver significant value that impacts product metrics. I asked two product managers from our AI infrastructure division to share their expert opinions.

Opinion #1: Alina Pavlova, Optimizing Processes with AI

AI, in itself, doesn’t create value; it’s a technology tailored for specific tasks. The true value of AI emerges when it addresses customer needs.

When considering AI integration, we ask: Can the need be met without using AI? Some tasks can indeed be solved without AI. However, in the future, we might integrate AI into standard solutions if necessary.

The following conditions should be met:

  1. Availability of Large Data Volumes
  2. Market Trends
  3. Routine and Repetitive Tasks

Upon deciding to implement, we always start with pilot projects to assess both the value of the AI solution and its feasibility.

I recall Sheldon Cooper’s phrase from The Big Bang Theory: “Everything is better with Bluetooth.” Nowadays, we don’t choose products based on Bluetooth availability, and the same applies to AI.

Opinion #2: Michal Karas, Implementing AI Features

Signs it’s time to implement AI:

  1. The product generates a multitude of data (client interactions/actions), whose value increases by establishing connections between them.
  2. Over 50% of this data is repetitive in type, logic, or scenarios.
  3. The product is large, and the data volume is continuously growing.
  4. The product aims to avoid exponential increases in data processing costs by utilizing solutions for typical tasks, thereby enhancing profitability.

OR

The product has rapidly expanded, and the data processing infrastructure hasn’t yet formed. It’s possible to set it up for efficient data handling (without human involvement, through algorithm training).

  1. The product’s core functions include:
    • Aggregating and Analyzing Large Data Volumes: Transitioning from mere data to trends, highlighting risks, and providing recommendations to mitigate them.
    • Facilitating Communication and Agreements: Built-in tools for verification, converting voice to text, transferring data to task trackers, and unified windows for procedures related to communication, such as verification, authentication, AI translators that vocalize contracts from “Chinese” or “legalese” to Russian, and generating voice summaries for lawyers or accountants.
    • Working with People: Tools for facilitation, gamification, emotion analysis, and summarization in areas like education, HR, real estate, psychology, legal practices, and group activities in large companies.
  2. The product can enhance efficiency within its core value proposition. For example:
    • Faster Customer Support: Quickly finding answers to client questions about the product, regulations, or document errors.
    • Real Estate Market Analysis: Identifying and purchasing properties faster than specialists for resale purposes.
    • Supply Chain Forecasting: Collecting data on product supply volumes to regions, forecasting prices, delivery speeds, product quality, and offering recommendations to clients within the existing product.
    • Financial Analysis: Evaluating factor proposals and suggesting application options to suppliers that are most likely to convert into received funds.

According to Gartner, supported by McKinsey, AI integration can bring value in three business areas:

  1. Effective Customer Service: Utilizing AI in support and sales to better understand customers, upsell, reduce churn, and accelerate purchases through improved need recognition.
  2. Resource Savings: Reducing the need for additional resources (primarily human or time) by having algorithms and agents handle part of the work:
    • Quickly gathering information within the company about process X (saving time for any employee, especially during onboarding).
    • Analyzing numerous calls to identify common issues and solutions.