The most revealing moment in corporate computing arrives just after something breaks. A laptop freezes. A video call stutters. An application develops the ambition of a glacier. The employee sees a spinning cursor; the help desk sees a ticket that says, with admirable precision, “computer slow.” By the time an agent responds, the evidence may be gone.
Lakeside Software has spent nearly three decades building a memory for that moment. Its flagship platform, SysTrack, collects performance and usage data from workplace endpoints, turns those readings into diagnoses, and gives IT teams tools to investigate or fix the cause. The company calls one historical view “Black Box,” a name borrowed from aviation. An agent can rewind a device to see the alarms, applications and system behavior around an incident instead of asking the user to recreate it.
That is the appealingly practical heart of Lakeside. It sells digital employee experience software, a category shortened to DEX and crowded with promises about productivity and happiness. But its product begins in less sentimental territory: CPU, memory, disk, boot time, application latency, software changes and thousands of other signals. Employee experience, in this telling, is partly an engineering problem.
The machine becomes a witness
Mike Schumacher founded Lakeside in 1997, when the original SysTrack focused on Citrix and multi-user computing environments. The vocabulary changed over the years - workspace analytics, end-user experience management, DEX - but the basic wager held: if IT could observe what happened on the device over time, it could replace anecdote with evidence.
SysTrack now sits across physical PCs, virtual desktops and other endpoint environments. Lakeside says its edge agent gathers more than 10,000 data points every 15 seconds. More than 1,400 sensors transform the stream into recognizable conditions, while dashboards and applications serve different jobs: the service desk needs a quick diagnosis; engineering needs patterns across the estate; asset managers need to know which licenses sit unused; executives want a view of cost, risk and productivity.
The SysTrack loop
Consider a recurring morning login delay. A conventional ticket records that someone waited. SysTrack can connect the wait to a particular process, policy, storage bottleneck or recent software change, then show how many devices share the condition. The problem changes from one employee's complaint into an estate-wide pattern. If the remedy is safe and repeatable, an automation can apply it without a technician visiting every machine.
“We instrument first, then we build dashboards, then we automate.”Dan Salinas, chief operating officer, describing the company's progression
What it sells - and who buys it
Lakeside is a private, enterprise software company headquartered in Boston, with roots and an office in Ann Arbor. Its supplied company record estimates about 230 employees, while LinkedIn places it in the 201-to-500 range. SysTrack is sold by quotation rather than a public menu. Deployment size, endpoint count and optional capabilities shape the contract, and the company sells both directly and through resellers, managed service providers and global integrators.
The buyers are usually large IT organizations with complicated estates: end-user computing teams, workplace engineers, help desks, asset managers and operations leaders. Public customer stories include investment manager PIMCO, legal-information company LexisNexis and Turkish glass manufacturer Şişecam, whose 22,000 employees make a tidy example of why scale matters. Lakeside also markets heavily to financial services, insurance, government, healthcare and manufacturing.
The basic platform covers collection, analytics, visualization and remediation. SysTrack Resolve gives service-desk staff live and historical device detail in one workflow. Sensors and automations detect conditions and run selected scripts. SysTrack AI, launched in late 2025 as a paid add-on, places a reasoning layer over the telemetry so agents and employees can ask questions, receive diagnoses and trigger governed actions. Integrations feed data into tools such as ServiceNow, Microsoft Intune, Power BI and Qualtrics.
In 2026 Lakeside pushed the pitch toward “reliability engineering.” SysTrack Reliability Engineering borrows the language of site reliability engineering - service-level objectives, error budgets and prevention - and applies it to the last mile between an application and the employee using it. That framing is smart. A business does not care that a server is green if the sales team's calls still fail.
What did it cost, and where is the return?
Without a public price list, the cleanest cost-and-payoff evidence comes from a 2022 Forrester Consulting study commissioned by Lakeside. Forrester modeled a composite enterprise and estimated $4.1 million in costs against $14.7 million in benefits over three years, a 259 percent return with payback in under six months. It is a model based on interviewed customers, not a promise every buyer should paste into a budget.
The savings logic is more useful than the headline percentage. Better diagnosis can reduce handling time and ticket volume. Usage evidence can find software licenses that nobody uses. Device performance can reveal which machines genuinely need replacement and which merely need a fix. Lakeside cites one international bank that identified $4.3 million in unused software and an insurer that avoided nearly 30 percent of help-desk tickets. Another published case described a 35 percent ticket reduction over a year.
These figures come from Lakeside materials and a commissioned study. They illustrate where value can appear; actual savings depend on estate size, contract terms, baseline performance and whether the customer changes operations.
What failed first: visibility
The first failure Lakeside attacks is rarely the processor or network. It is the gap between what the employee experiences and what IT can prove. Traditional monitoring often stops at infrastructure. Device management knows inventory and policy. Ticketing systems know what people reported. None necessarily preserves a fine-grained history of the endpoint at the moment of trouble.
That gap explains why the company evolved. SysTrack began as a system of record for end-user computing. The arrival of modern AI changed management's mind about where that record could go. Schumacher, who returned as CEO in April 2025, argues that AI is useful only when it reasons from trustworthy evidence. His company began machine-learning work in 2016, but the late-2025 launch of SysTrack AI connected that work to service-desk conversations, autonomous diagnostics and integrations with other agents.
“Plain facts first, decisions on top.”Mike Schumacher on Lakeside's evidence-first AI approach
The distinction matters because a confidently wrong support bot is not charming. It can waste an employee's time, corrupt a ticket or apply the wrong fix at scale. Lakeside's answer is to ground reasoning in continuous, first-party endpoint telemetry and expose how a diagnosis was reached. The strategy turns an old product asset - deep historical data - into the permission slip for a new AI product.
A crowded market with one sharp wedge
Lakeside competes with DEX specialists including Nexthink, ControlUp, TeamViewer's 1E and Riverbed Aternity. It also runs into broader suites. Microsoft Intune can manage devices; ServiceNow owns support workflows; observability vendors collect technical signals; internal teams have scripts. A buyer can reasonably ask why another agent and another console belong in the stack.
Lakeside's wedge
Deep endpoint history, broad sensor coverage, physical and virtual estate support, and diagnosis tied to remediation.
The buyer's burden
Prove that extra telemetry removes tickets and waste instead of creating one more dashboard for a thinly staffed team.
Lakeside differentiates on the breadth, frequency and history of its endpoint data. Its “Black Box” metaphor makes the advantage legible. The partner strategy makes it deployable: Microsoft and ServiceNow integrations put findings where work happens, Qualtrics can connect operational signals with sentiment, Intel adds device-health context, and firms such as Insight, SHI, SAIC, Peraton and Capgemini can carry the platform into complex accounts.
The weakness is the mirror image of the strength. More data is not automatically better judgment. A customer must decide which sensors matter, who owns each finding, which actions are safe and how success will be measured. Otherwise SysTrack risks becoming an exquisite weather station in an organization that never carries an umbrella.
What a reader can copy
You do not need 10,000 signals to borrow the company's operating idea. Start with one expensive, repeating complaint. Preserve enough history to reconstruct it. Connect technical symptoms to business impact. Create a tested remedy. Then measure whether the ticket returns. The order is important: automation belongs after evidence and diagnosis, not before.
The stealable playbook
- Choose a recurring failure with a visible cost, such as login delay or video-call instability.
- Instrument the whole employee path and retain history long enough to catch intermittent trouble.
- Translate raw readings into a condition that a support agent can understand and own.
- Test one bounded remediation, add permissions and keep a human escalation path.
- Measure tickets avoided, time saved and repeat failures, then expand only if the loop closes.
The approach works best in large, standardized environments where small improvements repeat across thousands of devices. It is less persuasive for a small company with a simple estate, or for an organization that cannot deploy endpoint collection. It can also stall where privacy rules are vague, teams distrust the data, automations lack change control, or nobody has authority to redesign the service desk.
There is a human condition, too. If managers use employee-experience telemetry as covert worker surveillance, they will destroy trust and muddy the purpose of the system. Good governance separates device and application health from judgments about individual effort, limits collection to a defensible purpose and explains the practice clearly.
Lakeside's most interesting achievement is not that it found an AI angle. Every enterprise vendor has found one. It is that the company can connect the angle to a long-maintained data asset and a specific operational sequence. Watch the endpoint. Remember the failure. Explain the cause. Automate only what is safe. For an IT department tired of learning about trouble from an irritated colleague, that is a grounded proposition.