AI Eng

From my observations, generative AI is sliding into the valley of disappointment. The peak of inflated expectations and wild excitement has already passed.

A call to a clinic: the operator cannot immediately make an appointment because the Internet is slow due to shelling, and promises to call back. They say goodbye, and the patient waits for a call. The operator has to record the lead in the CRM, look at the doctor’s schedule, and offer the patient an appointment. But he switches to new requests and forgets about the promise. In conditions of such multitasking and stress, even an attentive employee can make a mistake. For businesses, these are constant risks: lost customers, broken agreements, reputational damage. It’s not just about medicine — similar scenarios are repeated in other businesses. However, often it’s not the fault of the employee, but the result of «manual» processes. dev.ua, together with IPTel, a developer of systems for communications automation, tells what technologies help «insure» businesses against human errors and how it works in practice.

When the plan does not match reality, retail loses billions. According to POI and McKinsey estimates, up to 35% of all sales losses in FMCG are related to the human factor: 25% — poor implementation of standards, 8-10% — out-of-stock products (absence of goods on the shelf, — editor’s note). Empty shelves in stores and costs are a mirror of a systemic problem in the industry: there are agreements, but retail is still built on the human factor and manual processes that cannot be controlled on a large scale.

I have been working with remote developers for many years. This article will be useful for founders, CTOs, HR and developers working with AI and want to better understand the risks that have appeared in the market. I share the real experience of our company - without accusations and with clear conclusions.