By: Francesco Paola
24 May, 2021
In the first part of the blog, we highlighted several challenges facing today’s Service Desk and IT support ecosystems:
a) Alert Storms
b) Siloed monitoring solutions
c) Context and Correlation
d) Disparate data sources
e) Augmented Automation
f) Closing the Loop
Here, we discuss how a Machine Learning platform like Sainapse integrates with existing support infrastructure, preserving investment while at the same time addressing the identified challenges and enable the IT support organization to reduce the ticket resolution times, increase their first contact resolution rate, reduce cost per ticket, improve customer satisfaction and reduce total cost of ownership.
Alert storms are a very common challenge for IT
Support desks and is the one existing system monitoring
platforms and other systems of engagement that have had some success in
addressing. Essentially, it is
enabling an in-depth categorization of alerts and some machine learning
(ML) to ensure that the critical
alerts are handled by service desk and NOC reps first. The ability to
manage and triage alerts has become
table stakes in the monitoring ecosystem.
Sainapse enables alert triaging and management out of the box. Sainapse
learns how to segregate the
critical and actionable alerts based on two inputs – how it was done
manually in the past, and how alert
handling policies are defined for the NOC. Sainapse can also work
alongside existing tools that may be
implemented for this purpose, and power up downstream actions –
orchestrating remediation, and
resolution recommendation – that increase the efficacy and efficiency of
the human reps in the NOC.
Enterprises have also been investing in new and
improved technology platforms in order to keep up with
business needs and customer demand, but more often than not they do so
with departmental silos that
bring the promise of quick and easy deployment and integration across
the enterprise, but in reality, don’t
meet these lofty expectations. These point solutions, CRM, Service Desk,
APM, Knowledgebases, etc., are
independent silos that over time increase the level of technical debt.
And organizations are hesitant to
eliminate these platforms due to inertia and the level of investments
that have already been made.
Sainapse was architected to address the joint challenges of siloed
monitoring solutions (systems of
engagement) and disparate data sources. Understanding that the need to
preserve existing investment in
IT is predominant in the enterprise, Sainapse integrates with existing
support systems (APM, CRM, ITSM)
and data sources – knowledgebases, SOP, document repositories, and
third-party vendor data – without
the need to scrap prior investments or perform a massive systems
integration project.
Sainapse has also developed API integration with
some of the more common alerting and monitoring
platforms, such as AppDynamics, allowing it to receive the alerts from
the platform, research the
distributed data stores, query its ML engine for historical patterns,
determine the context of the alert,
correlate the information and provide a set of recommended resolutions
to the rep, reducing the time the
rep spends performing the research, enabling higher throughput of alert
management and ticket
resolution.
Furthermore, out of the box integrations with ITSM solutions such as
ServiceNow allow Sainapse to query
the CMDB to determine what if any configuration anomalies or changes may
exist, query the
knowledgebase or document repository to check the SOPs and Service
Agreements, query the ML engine to
determine context and correlate the events, and recommend a resolution
or a prioritized set of resolutions
for the rep to review and implement.
We touched upon augmented automation earlier,
where organizations have moved towards the mantra of
“automate everything” in their infrastructure, implementing
infrastructure as code, DevOps toolchains,
NLP, RPA, etc. Without an intelligent orchestration engine, the rep
still has to research the root cause of
the alert, research the SOPs and Service Agreements, find the right
automation script and manually trigger
the automation script.
Sainapse’s orchestration engine has the ability to determine which
automation scripts should be triggered
based on the context of the ticket and historical patterns (which its ML
engine is constantly updating),
trigger the automation scripts directly or if it is a sensitive
component of the system, provide
recommendations to the rep which automation scripts should be triggered.
One of the core strengths of Sainapse is how it
uses artificial intelligence combined with its machine
learning models to continually learn and improve its ability to make
more accurate recommendations.
Sainapse performs this function throughout the resolution process,
initially logging the recommendations
associated with a ticket and updating the recommendation in its ML
engine after the rep has either
confirmed the recommended resolution or changed the resolution to drive
a better outcome.
By closing the loop at multiple instances in the process, Sainapse
iteratively learns and enhances its ability
to make the right resolution recommendation the first time, allowing the
rep to focus on resolving issues as
opposed to spending countless hours researching for the correct
resolution.
Intelligent AIOps enabled monitoring platforms
and systems of orchestration need to be able to integrate
with the existing systems in the IT operations management (ITOM) chain
to enable a seamless workflow,
both upstream and downstream. Sainapse enables this capability out of
the box with a “no code, low code”
approach to configuration.
While current monitoring platforms and AIOps solutions primarily focus
on managing alert volumes and
ticket elimination, Sainapse focuses on enhancing the ticket resolution
process through ticket correlation
and determining context and providing prioritized resolution
recommendations, facilitating the work of the
IT support team and driving towards the achievement of the TCO, Revenue
and Customer Satisfaction
benefits.
For example, if a resource is low on memory or a VM close to capacity,
i.e., near or at its efficiency
threshold, Sainapse receives the alert from the monitoring platform,
confirms that the said resource is
supporting a critical service, and then either directly kicks off a
runbook via an automation workflow
(CI/CD) platform, e.g., Ansible that automatically allocates incremental
resources, or recommends the
automation path to the rep. Once the new resource is allocated, Sainapse
and the monitoring platform are
notified of the new resource allocation, the initial alert is closed and
Sainapse logs the resolution,
correlating it to the alert as it continuously enhances its ML
model.
Sainapse leverages existing IT support infrastructure investments by
seamlessly integrating into common
CRM, APM, ITSM platforms and diverse knowledgebases (e.g., Sharepoint,
Dropbox etc.), preserving IT
investment. As the system of orchestration, Sainapse facilitates
augmented automation so that the support
reps can focus on executing the fix as opposed to spending time
performing research.
Finally, Sainapse iteratively learns throughout the process, closing the
loop so that it institutionalizes the
learnings and continually optimizes its recommendations.
A system that is Sainapse-enabled increases Service Desk and IT Support
teams’ operational efficiencies by
leveraging distributed data, correlating incidents and service requests,
extracting insights about the
underlying systems, monitoring operational and usage statistics, and
proactively recommending resolution
options or automatically solving application performance problems.