Feature
Information technology (IT) teams are embracing artificial intelligence (AI), but how they’re using the technology depends on the nature of the company and the type of work they’re doing.
IT professionals are among the earliest adopters of AI, with 95% using it weekly — which is significantly more than most other workers, according to a report by Freshworks.
Here, a group of AI leaders share exactly how they’re seeing AI used by IT teams to assist organizations.
Overhauling Legacy Systems
Robin Patra — head of data, platform, product and engineering at ARCO Construction and previously at BlackRock — said AI is “reshaping IT leadership, and enterprise architects are at the forefront of this transformation.”
For example, an IT leader at an enterprise financial services firm "leveraged AI to overhaul legacy systems integration,” Patra said.
Traditionally, this has been a painstaking and error-prone process. Getting it done was frustrating for human teams.
By deploying AI-powered natural language processing (NLP) tools, they were able to map data fields across incompatible systems, identify inconsistencies and automate most of the integration process, according to Patra. This reduced the project timeline and saved the company millions in operational costs while improving data accuracy.
Patra said the project had its challenges, though. Mostly, this was in the form of resistance from mid-level IT staff who were concerned about their job security and the reliability of the AI. Education and transparency about what the AI was doing helped. Once the team understood the AI was a tool, and not hired to replace them, the team became more collaborative.
Classifying and Labeling Images
Carlos Meléndez — VP of operations of Wovenware, a software development firm — said his IT team tapped the speed and intelligence of AI to classify and label a vast number of images, a task that is slow and burdensome for humans.
“AI is helping to automate the reporting and analyzing of satellite images so that in addition to images, we’re providing clients with real insights about what they’re seeing,” Meléndez said.
To accomplish this work, the AI uses computer vision to identify objects in the images.
“This type of automated object identification helps us derive deeper insights about what the images are telling us,” Meléndez said.
The technology saves time and makes a time-consuming service possible, but it does have challenges.
“It’s important to remember that it’s really always about the data,” Meléndez said. “For AI to work effectively, the data must be clean, classified and relevant.
“Another lesson is that an AI project is never really done. Unlike software development projects that have a beginning and an end, AI algorithms must always be trained on new data to be current.”
Related Article: 10 Top AI IT Products
Automating Data Collection
Dima Eremin, the CEO of BluedotHQ, a note-taking company, said his IT team uses AI to automate data collection and analysis.
“This is the most common use of AI, in my opinion,” Eremin said.
AI solves a big problem at the company: It can quickly analyze large amounts of to help drive the company’s goals and decisions.
“Analyzing data yourself is a lot of work that is also very vulnerable to human error, which AI eliminates,” Eremin said.
The company’s platform uses a natural language processing model, which supports data analysis at the organization.
A challenge Eremin has seen other companies face is the difficulty of integrating AI into existing systems, because they end up with an AI tool that doesn’t support the tools they’re already using.
“I suggest checking that before you invest in AI adoption,” Eremin said. “Additionally, your models need to be trained and reviewed regularly in order to prevent bias.”
Finding Errors
Mike Finley, CTO of AnswerRocket, a data analytics company, said when it comes to off-the-shelf generative AI tools, two of the primary adopted IT use cases are for improving the quality of code and documents.
Tools like Cursor.ai can deliver nearly instantaneous developer productivity of 30% or more by pointing out errors, suggesting next steps and being "a partner on the shoulder,” Finley said.
As AI technologies evolve, standardized solutions are beginning to emerge, such as workflow tools that look at specific data and identify patterns, like finding errors in invoicing or conditions in contracts, according to Finley.
Webinar
Oct
16
Agentic AI Playbook: Real-World Customer Service Use Cases You Can Deploy Now
Boost self-service by 30% and slash call volume by 63% with agentic AI.
Register
Webinar
Oct
22
Beyond Storage: Smarter Content, Bigger Impact with DAM + AI
Discover how the DAM + AI duo makes content smarter, stronger and more accessible.
Register
Webinar
Nov
6
Fix the Content Bottleneck: Build a Better WebOps Strategy
Content stalled? Dev overloaded? You’re not the only one. Learn how streamlined WebOps bridges the publishing gap.
Register
Webinar
On demand
CMS Briefing: A Live Look at What’s Next in AI-Driven Platforms
Learn how leading organizations are using AI‑driven tools to publish faster, personalize smarter and stay secure.
Watch Now
Webinar
On demand
Ready or Not: How Data-First Organizations Are Unlocking Agentforce Potential
Learn how to cut through the noise, activate Agentforce and build a Salesforce AI strategy that actually delivers.
Watch Now
Webinar
On demand
AI in Customer Service: Faster Resolutions, Happier Customers
Don’t let rising demand burn out your team. See how to build a smarter, more resilient support org.
Watch Now
“These are things people can do, given instructions and access to information, but GenAI can do it faster, looking for smaller pickups that really add up,” Finley said.
There are still "hazards,” he said, when it comes to using AI. Investing in a bespoke solution can help navigate many of them.
At the same time, Finley stressed that IT teams shouldn't create errors when they “back a language model into a corner.”
“These models are trained to please humans,” Finley said. “If the only way to finish a job is to hallucinate a fact or two, the model will do just that.
“As we empower these models with the ability to take actions, not just teletype text back to a screen, hallucinations become very expensive. Always provide it with a way to ask for help.”
Related Article: 5 AI Case Studies in IT
