Getty Images/iStockphoto

IBM gives glimpse of HashiCorp plans with AIOps preview

The idea that IBM's $6.5 billion deal for HashiCorp will change its open source stance has lost steam given plans to unite it with other proprietary tools under IBM Concert.

IBM unveiled plans to ship a new AIOps product next month that will tie in tools from HashiCorp, along with Red Hat Ansible, IBM Turbonomic, IBM Instana and third-party vendors.

The product, IBM Concert, will tie together these tools using the Watsonx generative AI platform to create a comprehensive view of enterprise applications, issue natural language recommendations and orchestrate automated responses to events, according to Kareem Yusuf, senior vice president of product management and growth at IBM Software, during a press briefing Monday.

"Think of it ... as essentially a nerve center, a generative AI-powered brain that is making sense of application environments," Yusuf said. "It uses its LLM-backed data model to correlate and derive insights from all this data to give an accurate view on the application landscape ... and then that allows us to begin to drive various scenarios."

Initial scenarios envisioned for IBM Concert include assessing the effects of security vulnerabilities on a company's applications and quickly triaging those most at risk, he said. IBM Concert can then kick off AIOps workflows using the integrated tools.

Yusuf declined to comment on whether IBM will change HashiCorp's controversial Business Source License, saying IBM's planned $6.5 billion acquisition, disclosed last month, has not yet closed. Initially, industry watchers speculated that IBM's support for open source forks of HashiCorp's Terraform in OpenTofu and Vault in OpenBao might mean it would reverse that course.

But besides Ansible, the other tools mentioned as part of IBM Concert are all proprietary, which makes such a change seem increasingly unlikely, said Rob Strechay, lead analyst at enterprise tech media company TheCube.

"That licensing change is not going to happen. ... [HashiCorp] already did the hard work [with that]," Strechay said. "IBM will stay close enough to OpenTofu to monitor and contribute nondifferentiated code."

Deeper ties between Terraform and Ansible planned

Yusuf also said this week that IBM will directly integrate HashiCorp tools with parts of its Red Hat portfolio, specifically HashiCorp Terraform and the Ansible Automation Platform.

As you can imagine, both Ansible and Terraform from HashiCorp represent triggering and information-pulling environments that fit into [IBM Concert]. We see synergistic value in integrated product capabilities from HashiCorp Terraform and Ansible, as the two most obvious ones.
Kareem YusufSenior vice president of product management and growth, IBM Software

"As you can imagine, both Ansible and Terraform from HashiCorp represent triggering and information-pulling environments that fit into [IBM Concert]," Yusuf said. "We see synergistic value in integrated product capabilities from HashiCorp Terraform and Ansible, as the two most obvious ones."

HashiCorp Terraform and Red Hat Ansible are both infrastructure-as-code tools with some overlapping features, but have different strengths that mean they are often used together by Ansible Automation Platform customers. Terraform tends to be favored for initial raw infrastructure setup, and Ansible for ongoing configuration.

Terraform and HashiCorp's Vault secrets management product are already used together, but IT pros said during Red Hat Summit and AnsibleFest earlier this month that there would be benefits to more closely aligning product roadmaps between the two once they're under the IBM umbrella.

"You can definitely use them together today, but there's always room for improvement in terms of seamless integration," said Yamandú Tellechea, automation lead at Madrid-based multinational insurance company Mutua Madrileña, in an online interview during the Red Hat conference. "Currently, you need to set up and manage separate configurations for each tool, which can sometimes lead to inconsistencies or duplicated efforts. What I'd like to see is a more unified interface or workflow that simplifies how Ansible and HashiCorp tools interact."

The most straightforward way to do this would be integration through the Event-Driven Ansible tool added to the Ansible Automation Platform last year, said Jim Mercer, an analyst at IDC, in an interview during Red Hat Summit.

"A simple one would be using EDA as a means of [offering] event-driven Terraform," Mercer said. "I am sure there will be [other] new use cases where these [tools] can be better integrated to address cloud and AI automation."

IBM's long -- at times, rocky -- AI history

There are some surface similarities between IBM Concert and other IBM AIOps products, such as its now-former QRadar security product line, sold to Palo Alto Networks last week. QRadar's security orchestration, automation and response module also kicked off automated playbooks for threat response based on AI risk analysis.

QRadar was specifically for security incident response, while IBM Concert is focused on security and other types of IT automation for applications. At first glance, it overlaps with the company's existing IBM Cloud Pak for AIOps, which also "leverages AI to streamline IT operations, increase uptime, improve efficiency, and reduce costs ... [and] supports Operations teams to deliver optimum performance and assurance by ... providing real-time insights into IT systems [and] augmenting incident management with AI," according to IBM documentation. Version 4.4.1 of this Cloud Pak added integration with OpenTofu.

But IBM Concert uses machine learning and generative AI along with "a novel approach to data modeling that accelerates the ability to obtain actionable insights from data aggregated across IT tooling silos with a Watsonx conversational experience," according to a company spokesperson in an email to TechTarget Editorial. "Additionally, Concert uses generative AI to help with application compliance standards ... [and] give developers recommendations or code snippets to accelerate the resolution of application vulnerabilities."

Still, IBM's overall history with AI products -- which includes its notorious and now-scrapped Watson Health division -- came up in questions during the press briefing. Company officials said they've learned from those experiences and used them to inform the development of Watsonx. There's also a growing belief among industry experts and IT pros outside IBM that generative AI might help boost the effectiveness of AIOps tools based on other kinds of machine learning.

Graphic showing survey results on the impact of AIOps.
A 2023 survey from TechTarget's Enterprise Strategy Group found that enterprise AIOps has had limited effectiveness to date. Some industry experts predict that generative AI might help.

However, many enterprises are still struggling to get generative AI into production, said Mohamed Ali, senior vice president and COO of IBM Consulting, during the briefing session.

"About 42% of [IBM Consulting] clients with greater than 1,000 employees have started to take their pilots and put them into production for these GenAI use cases, [while] 40% are what we call stuck in the sandbox -- still experimenting and not quite sure how to take it to production," he said. "There are security issues, they have PII [personally identifiable information] issues, bias issues, code issues, intellectual property issues. ... In IBM Consulting, we've been working with both categories of these clients."

If IBM and Red Hat aren't careful, pushing generative AI offerings that are not ready could go the way of Watson Health, said Larry Carvalho, an independent analyst at RobustCloud, a cloud advisory firm.

"Promised huge ROI, organizations are implementing generative AI products, including Ansible Lightspeed, without ensuring data quality and adequate testing," he said. "Customers should evaluate the nascent technology before taking the leap based on vendor hype."

Beth Pariseau, senior news writer for TechTarget Editorial, is an award-winning veteran of IT journalism covering DevOps. Have a tip? Email her or reach out @PariseauTT.

Dig Deeper on Systems automation and orchestration

Software Quality
App Architecture
Cloud Computing
Data Center