Shailesh Manjrekar is the Chief Marketing Officer at CloudFabrix, the inventor of Robotic Data Automation Fabric and an AIOps Leader.

Generative AI (GenAI) is getting more integrated into solutions we use individually or as part of organizations, and AIOps and observability are no different.

GenAI is expected to catapult this market by making AIOps and observability more accessible and efficient. However, there are challenges that come with using GenAI without working on its underlying data. Let’s explore.

What Is Generative AI For AIOps?

Generative AI produces new content. It isn’t new, though. GenAI’s roots go back to machine learning. Most recently, in 2014, Ian Goodfellow and colleagues introduced generative adversarial networks, or GANs, a kind of ML algorithm that could create authentic videos, images and audio of real people—the GenAI we know today.

Transformers and large language models (LLMs) fuel GenAI applications. I highly recommend this insightful read on the subject to learn more about the evolution of GenAI.

Today, we’re reviewing how GenAI augments IT operations: AIOps and observability. Data and AI represent power and opportunity in the digital age. GenAI democratizes data and AIOps by making them easy to understand, access and work with.

Operations teams are continuously inundated with data. AIOps and observability have become the bare bones of operations, and GenAI holds a new promise, albeit with a caveat.

Allow me to make a case for GenAI in AIOps and observability. To do that, let’s understand what goes into a successful GenAI implementation.

Context-Aware Conversational Queries With Local LLMs And Data Fabric

The effectiveness of a co-pilot or GenAI assistant depends on how well it can understand the context of a conversational query (the input in human conversational language) and yield relevant results.

An example of this is asking, “Who discovered gravity?” followed by “How was it discovered?” and “What is its significance in science today?” The follow-up questions require the algorithm to maintain and use the context of the first question. It needs to understand that we’re still talking about gravity even when we don’t explicitly repeat the context.

In AIOps and observability, organizations want GenAI to answer/perform the following:

• Create a composable dashboard for syslog data sets.

• Create a new RDA (robotic data automation) pipeline to create ServiceNow tickets.

• What data sets do I have in my RDA platform?

• Explain network outage impact upstream.

To be this specific, GenAI needs the right data within the appropriate context. Data fabric allows real-time discovery of the environment and inclusion of all the data sets. Then, it enriches the data with automated topology discovery or external feeds.

RDAF (robotic data automation fabric) builds the context for a conversational query. Low-code/no-code pipelines enable rapid code deployment to build this context and enrichment.

GenAI implementations where LLMs are trained using only OpenAI frameworks lack relevance. Since the algorithm builds on publicly available generic data, it fails to be relevant to your data, environment and pipeline.

Context Awareness Unlocks GenAI’s Highest Potential For IT Operations

How do we build this context? We use a local knowledge corpus. We embed composable low-code/no-code bots, knowledge base articles, metadata, service blueprints and normalized enriched data with context into a vector database.

A local LLM can understand the local context, turn a conversational query into a context-aware prompt and send it to the OpenAI service. For each query, a semantic search using the Langchain framework adds the context and sends it to OpenAI. Now, the OpenAI service yields extremely relevant information or carries out tasks based on the environment.

So, how can all this contextual information be sent along with the prompt without exhausting the token limit on OpenAI prompts?

A retrieval system with contextual document compression helps reduce the prompt size while maintaining context information.

GenAI alone isn’t enough. Context, and by extension, the underlying data are critical.

Here are a few challenges that come with GenAI for AIOps and observability:

• Ensuring The Accuracy And Predictability Of LLMs: Training an LLM on public data sets may introduce errors and misinformation. It’s best to train LLMs on custom data sets using semantic search and RDAF to build a local knowledge corpus.
• Ensuring Privacy And Governance Over Custom Data Sets: User policies are key to achieving data governance.
• Working With The Token Limitation On OpenAI’s GPT: Data compression is important while maintaining the RDAF-generated context with the local LLM for accurate, context-aware prompts.
• Asking A Recurring Question: Using locally generated context, with RDAF and local LLM to feed the prompt, enables accurate responses to follow-up prompts.

GenAI, Data Fabric, Data Automation And No-Code Pipelines: The Winning Recipe

There are usually two inhibitors to AIOps:

1. The Right Data: Getting the data in order requires a data automation framework. Data constitutes the raw material (for training), as well as the dressing (context) in the recipe for GenAI.

2. Democratization: AIOps and observability must be achieved and handled org-wide. GenAI makes this possible, but only with the right context.

This idea is close to the data value gap I often talk about, which creates a gap between data and the derivation of value from it. The data value gap consists of:

• Data Quality: 80% of the time is spent on getting the data right.

• Data Silos: Data is dispersed across the endpoints, data centers and multicloud.

• The Skills Gap

The unique combination of RDAF, data automation, no-code pipelines and GenAI works on all three challenges to close the data value gap.

A low-code environment enables rapid code development using bots. This means a GenAI implementation can be achieved within weeks instead of months for an organization. RDAF helps build the context for conversational queries by making real-time data accessible and available.

Data automation helps massage and enrich the data before using it for a local LLM. Finally, GenAI works its magic and unlocks the potential for AIOps and observability.

GenAI holds high promise for AIOps and observability, and this market is expected to catapult. RDAF, no-code pipelines and data automation are paramount for GenAI to be contextual and accurate in the AIOps use case.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?

Any business hoping to succeed in today’s competitive landscape understands the need to put digital transformation at the top of its priority list. 

Digital technology now infiltrates almost every aspect of conducting business, from how customers interact with organizations, how employees do their jobs and how the internal processes work, and the global pandemic and shift towards remote working only accelerated this trend. Put simply, those which do not adapt to the digital world are highly unlikely to survive. 

“Both customers and employees expect to have a contextual experience throughout their journey with the company,” says Prabhu Karunakaran [PK], founder and CEO of digital transformation firm Exterprise. “To achieve such seamless experiences in a hybrid work environment, organizations are creating more technology than ever before, and are becoming digital solution companies in the process. They are quickly realizing that they need to unite their formerly siloed information technology, business, HR, legal and marketing departments into fast-moving, cross-disciplinary digital teams.” 

But for established businesses, this often means transforming from rigid legacy systems to flexible digital platforms, which can prove challenging even for Fortune 500 companies. A lack of historic investment in technology has been compounded by budget cuts during the pandemic, meaning many still rely on antiquated hardware and software installed decades previously. 

This is not the only factor holding businesses back. Other common issues include organizational silos where decision-making can be split across functions; a lack of digital skills; a culture which is resistant to change; and a lack of clear strategy and communication across departments. 

For many businesses, the answer is to turn to one unified platform that can bring productivity improvements through digital transformation which can help employees, customers, and assets. “Taking good care of these three core imperatives is critical to profitable revenues and organizational growth,” says PK. “Productive and happy employees will deliver enhanced customer experiences, resulting in more revenues. Efficient monitoring of assets including hardware, software and enterprise assets will create operational excellence.” 

One such offering is ServiceNow’s NOW Platform, which deploys enterprise-scale solutions to organizations in several industries, improving efficiency for employees, customers, and assets. Its IT Service Management solution, for instance, automates core service processes, and makes it easier for those working in IT to resolve issues faster, while its IT Operations Management and Asset Management package proactively monitors the performance of assets, helping to prevent outages in the first place. 

The Customer Service Management element ensures agents have the information they need to quickly resolve issues, including proactively identifying potential problems before they become apparent. Businesses can also benefit from automated responses to common customer questions, increasing the efficiency of employees. Staff working in the field can benefit from the same functionality through the Field Service Management solution. 

From an employee perspective, businesses can deliver a streamlined service experience with intelligent workflows, starting with efficiently onboarding new hires. Employees themselves can find the information they need to self-service their own HR needs, while the Workplace Service Delivery solution helps the day-to-day work run more smoothly, boosting productivity. A exact IDC report suggests agile organizations retain employees at a 34% higher rate than competitors, helping them meet the demand for skills in the current war for talent. 

With teams in the US, India, and Central America, Exterprise is a certified pure-play ServiceNow implementation partner, with skilled resources operating in different time zones, and clients in sectors including financial services, insurance, healthcare, telecom, and government. Exterprise helps businesses on their transformation journey, starting by determining that the organization is fully committed to the journey, and helping to provide advisory and consultative services to demonstrate the business case and identify any challenges that will need to be overcome. 

“Partners are critical to ServiceNow’s growth strategy, on our journey to $16B in revenue and beyond,” said David Parsons, Senior Vice President of Global Alliances and Partner Ecosystem at ServiceNow. “Enabling great customer and employee experiences to faster time to value, we remain committed to engaging our partner ecosystem to innovate and co-create end-to-end solutions that solve today’s most pressing business challenges.” 

Exterprise works closely with clients to launch the digital transformation journey, identifying automation and integration opportunities and creating a strategic roadmap for the different departments, as well as training staff on how to use the platform. 

Exterprise helps in identifying key metrics around digital standards – including mean time to resolution (MTTR) for issues, customer satisfaction (CSAT), net promoter score (NPS) and first call resolution (FCR) for customer service. This ensures organizations realize the full value, both financially and in other, less tangible ways, of their investment. Exterprise is SOC2, ISO 27001 and HIPAA certified, which can deliver customers an added layer of reassurance. 

Ongoing reviews ensure businesses benefit from and understand any new functionality to keep enhancing the employee and customer experience, with ServiceNow introducing two major version upgrades each year. This ensures customers remain at the forefront of digital transformation, turning them from digital laggards to leaders. 

One example of a business that has benefited from running several business functions on the ServiceNow platform is pipeline services and equipment company T.D. Williamson. “Exterprise has proved to be an effective partner in not only implementing ServiceNow, but also in providing consultative and advisory support to help us automate core business processes,” says Greg Rice, Director, Global Applications. 

PK believes those who take the step to become a digital business can expect to benefit in several ways, including enhanced customer and employee experience, lower employee attrition rates, productivity benefits through process automation, and ultimately increased – and sustainable – revenue and profit. 

“The time to act is now,” concludes PK. “Those who do so can put in place the foundations that will deliver ongoing success in years to come, increasing efficiency and ensuring they have satisfied customers and employees. That’s the recipe for success.”

To find out more about how Exterprise could help your business, visit exterprise.us/servicenow/

To contact us, click here or email us at contact@exterprise.us

In the first six months following the Russian invasion last February, some 13 million people were displaced. It’s Europe’s largest refugee crisis since World War II. To respond to this vast and urgent demand for help, the U.S. government created a new visa program, Uniting for Ukraine, in April.

But the program requires something most refugees don’t have: a sponsor in the United States. How does a Ukrainian refugee find a sponsor in a country of 330 million strangers—especially if they’re 5,000 miles away, hiding in a bomb shelter? Especially if they don’t speak any English? And if you’re an American volunteer, how can you safely make contact with a Ukrainian who’s seeking your help?

With these questions in mind, Welcome.US, a non-profit initiative working to expand capacity to welcome newcomers to the United States through the increased and direct involvement of Americans, approached Goldman Sachs to propose they collaborate on a similar project for Ukrainians. A team of engineers at Goldman, led by a Ukrainian American refugee and a Russian American refugee, sought consulting advice on the project from Infosys and enlisted software company ServiceNow to build the website. From the final design to the day it went live, making Welcome Connect took just two and a half weeks—thanks to ServiceNow’s low-code platform, which can be swiftly customized, scaled, and implemented.

Welcome Connect bypasses barriers of distance, red tape, and language by matching Ukrainian refugees with Americans seeking to sponsor them. Within days of the site’s launch, some 5,000 Ukrainians were already communicating on it with prospective American sponsors, using its instantaneous text-message-translation feature.

While tens of thousands of Ukrainians have been granted parole in the United States, there are still thousands more in need of refuge as the war wages on with no end in sight. This is the story of two families: the Yeremenkos, of Kharkiv, Ukraine, and the Teiglands, of Kansas City, Missouri.

The University of Dayton School of Business Administration (UD SBA) is accredited by AACSB International - The Association to Advance Collegiate Schools of Business. Additional specialized AACSB International accreditation is held by the UD SBA Accounting Program.

Meeting AACSB standards is a measure of highest excellence and standards in curriculum, faculty, staff, facilities, and student body. To achieve accreditation, business and accounting programs must satisfy the expectations of a wide range of quality standards relating to the strategic management of resources, interactions of faculty and staff with students in the educational process, and achievement of learning goals in degree programs.

As part of our accreditation, we maintain a strong Assurance of Learning program that allows us to use well-documented, systematic processes for determining and achieving our degree program learning goals. Learning Goals are the educational expectations for our degree programs. They specify the intellectual and behavioral competencies our programs will instill in graduates.

Undergraduate Program Learning Goals and Objectives

(Approved by SBA Faculty October 2016)

Learning Goal 1: Our students will be knowledgeable of core business concepts.

• Objective 1: Our students will demonstrate knowledge of core concepts in business disciplines.
• Objective 2: Our students will articulate awareness of how an intercultural world affects business decision-making.

Learning Goal 2: Our students will be ethical decision-makers.

• Objective 1: Our students will apply an ethical decision-making model such as Catholic Social Teaching or sustainability to offer potential solutions to ethical dilemmas that arise in business practice.

Learning Goal 3: Our students will be entrepreneurial with their approach to problem-solving.

• Objective 1: Our students will develop potential solutions to business or societal problems by identifying opportunities that draw upon a multi-disciplinary creative approach.

Learning Goal 4: Our students will be proficient in applying practical business skills.

• Objective 1: Our students will serve as effective group members who contribute to team success.
• Objective 2: Our students will demonstrate ability to use technology to retrieve, structure, or analyze data to conduct research and make decisions.
• Objective 3: Our students will demonstrate ability to communicate effectively.

MBA Graduate Program Learning Goals

Learning Goal 1: Students will have an ability to address business problems based on application of effective analysis/use of decision-making tools.

Learning Goal 2: Students will be able to integrate and apply knowledge from various perspectives and viewpoints to business strategy, problems and practice.

Learning Goal 3: Students will be able to effectively use skills appropriate to managing and leading.

Learning Goal 4: Students will demonstrate understanding of importance & scope of ethical behavior in personal/professional context.

MBAN Graduate Program Learning Goals

Learning Goal 1: Students develop Understanding and Skills in Data Management, able to design and manage organizational data acquisition, storage, retrieval, and display technologies to meet business objectives.

Learning Goal 2: Students develop Skills in Analytics, abilities in using modeling and analysis methodologies to yield insights supporting effective organizational operations and decision making.

Learning Goal 3: Students develop Skills in Organizational Problem Solving, effectively interacting with users and decision-makers in problem diagnosis, alternative generation, analysis, solution-finding, and solution implementation.

Learning Goal 4: Students develop Skills in Communication, effectively communicating data and analytics results and insights within the organization.

Learning Goal 5: Students develop an Understanding of Ethical Issues, abilities to evaluate and apply ethical principles in using data and analytics in support of organizational decision-making.

Master of Professional Accountancy Graduate Program Learning Goals

Learning Goal 1: Our students will be ethical accounting and business professionals.

Learning Goal 2: Our students will have accounting (as broadly defined) research competence.

Learning Goal 3: Our students will be critical thinkers.

Learning Goal 4: Our students will have data analytics skills.

Master of Finance Graduate Program Learning Goals

Learning Goal 1: Our students will develop in-depth knowledge of the fundamental financial concepts.

Learning Goal 2: Our students will understand the importance and scope of ethical behaviors in both a personal and professional context.

Learning Goal 3: Our students will think critically in devising financial solutions.

Learning Goal 4: Our students will effectively conduct quantitative analysis.

Description of Learning Communities

A Learning Community (LC) provides opportunities for students to:

• Attend two to three classes with the same small group of students;
• Live in the same residence hall with members of the same community;
• Get involved, meet people, and make friends;
• Attend out-of-class programs and go on exciting trips, many of which relate directly back to what is being discussed in the classroom;
• Learn about helpful resources available at Purdue University; and,
• Become involved in the Greater Lafayette community.

At Purdue, a Learning Community is either:

• A group of 20-30 first-year students who take 2 or 3 of the same courses together;
• A group of first-year students who share a common academic interest and live in the same residence hall; or,
• A group of first-year students who take part in both of the activities listed above.

Two new studies shed light on the promise—and challenges—of learning recovery strategies that schools adopted in the wake of the pandemic.

The key theme from the research is as much a truism as it is a prescription: Getting the implementation right really matters. And it’s also difficult.

One working paper, from researchers at the American Institutes for Research, Harvard University, and the assessment provider NWEA, analyzed eight districts’ summer 2022 learning programs. Researchers found that the programs had a small but positive effect on student math scores, and no effect on reading.

Another working paper, from researchers at Vanderbilt University and Stanford University, examined tutoring programs—specifically, what conditions led to successful implementation. They identified some commonalities: Programs that functioned well had the support of school leaders and a structured process for recruiting tutors and managing logistics.

Together, the two studies underscore that academic recovery strategies require sustained, coordinated effort, said Susanna Loeb, a professor in the Graduate School of Education at Stanford, and one of the authors on the tutoring study.

“Anything that you’re going to do in schools, whether it is high-impact tutoring or a summer program, in order to have the kinds of impact that we need to get students engaged in school and learning at a faster pace, … you’re going to need a lot of effort into implementation,” Loeb said.

Giving students many and varied opportunities for extra learning is important, too, said Emily Morton, a researcher at AIR’s National Center for Analysis of Longitudinal Data in Education Research and one of the authors on the summer learning study.

The summer programs, which included time for academic instruction and enrichment activities, helped students make “marginal progress”—progress that is meaningful, she said.

“This is some of the very first good news we’re getting on any COVID recovery,” Morton said. Still, she added, “layering supports is going to be essential. Students can’t just get summer school.”

Districts have used federal COVID relief funds to support summer programs, tutoring initiatives, and other academic recovery plans—money that runs out in September 2024.

But maintaining multiple options for students requires funding and resources, Morton said. “To be able to do that, they’re going to need support from states and even from the federal government.”

Summer programs show small gains in math

The summer school study focused on summer learning programs in eight school districts. All of these districts served higher percentages of Black and Hispanic students and higher percentages of students eligible for free- and reduced-price lunch than the national average.

The researchers compared students who were eligible to attend summer school and did attend to students who were eligible but did not attend. They controlled for students’ characteristics, including race and free- and reduced-price lunch status, and their spring 2022 achievement levels.

Students who attended summer programs scored higher on an assessment of their math skills than students who did not. The researchers found a small, but positive, effect of summer learning on math achievement—0.03 of a standard deviation, which they estimated would mitigate about 2-3 percent of the total learning loss experienced in those districts. Most of this improvement was driven by upper elementary students.

The researchers found no effect for summer school in reading.

The programs did help students get extra learning time—the researchers found that each day of summer school produced academic gains roughly equivalent to each day of a regular school year. But these gains only made a tiny dent in academic recovery writ large.

Programs in the study were between 15-20 days long. On average, only 25 percent of targeted students participated. “There’s an opportunity here to scale summer school more widely, and potentially to make it longer,” said Morton.

These kinds of changes would bring programs more in line with summer learning best practices, said Jodi Grant, the executive director of the Afterschool Alliance, an advocacy organization. Grant was not involved in the study.

“Even with intense support, it’s really hard to move the needle over a short period of time,” she said.

Grant also cited research-backed recommendations that summer learning should include opportunities for enrichment activities—not only academics. Doing so can make summer programs more attractive; they shouldn’t have to feel like a “punishment,” she said.

Conditions for effective tutoring programs

Designing an effective program is the first step. Implementing that program presents a whole new set of challenges.

This piece of the puzzle—implementation—is the focus on the working paper on tutoring.

“We know this intervention is really, really effective,” said Loeb. “But what hasn’t gotten nearly as much attention is that it’s not that easy to implement.”

Researchers reviewed 33 articles published since 2000. They found strong implementation relied in large part on the buy-in of principals and school leaders.

“Education leaders served as gatekeepers of student and staff time, school space, and data/documentation on students’ skills and instructional needs,” the researchers wrote.

Having a dedicated and informed program manager was important, too, Loeb said. “You really need someone at the district level who is kind of organizing the approach, making sure that it aligns with what we can see from the results are the most promising practices,” she said.

Program managers need to be compensated for the work they’re doing, she added. That money could come from new funding, or districts could reprioritize funds. “But you definitely need somebody who has paid time to do it, and who is really responsible [for] it,” she said. “You can’t just add to people’s work without taking something away.”

Meeting conditions like these has been challenging for school districts.

The National Center for Education Statistics’ School Pulse Survey, an ongoing study of how schools are changing since the pandemic, polled school leaders in December 2022.

About a third of school leaders said they didn’t have the capacity to provide high-dosage tutoring to students who needed it. Two out of five schools said they couldn’t find time in their regular schedule for tutoring.

Still, Loeb said, better implementation is “not undoable.”

“Because it’s not easy,” she said, “we should be putting effort into thinking about the implementation process and really working on doing the implementation well.”