Thomas Carnevale

Speaking, interviews, and the questions worth asking

I speak to industry audiences, school and government leaders, executives, and newsrooms about the gap between security promises and operational performance. A useful talk should leave an audience with questions they can take back to a site or procurement meeting. My own material draws on developing panoramic video software, leading a manufacturer, real deployments, and independent evaluation of products in the field.

My topics can be shaped for a technical conference, executive briefing, school safety event, public-sector panel, webinar, or newsroom deadline. I can address how systems are designed and governed without pretending that one product or manufacturer solves every problem.

I have spoken at industry events including ISC West, ISC East, Intersec Dubai, and ISCEX. A conference appearance is a chance to explain what worked, where deployment gets complicated, and what the buyer should verify. If you are planning a panel, interview, or technical conversation, send the topic and timing.

These are sample talk directions. Share your audience and the decision they face, and we can shape the content accordingly.

What I told WGN

Buying cameras, alarms, and card readers is only the start. We need to know whether the pieces work together, whether staff know how to respond, and whether the result makes the facility safer. The interview gives a more direct version of that argument in my own words.

Those questions change how I evaluate a proposal: who receives an alert, who can confirm what happened, and what action can they take next? A product demo rarely answers all three.

In the interview, I explain how disconnected systems leave gaps (around 0:51), how a physical key differs from role-based access (around 1:06), and why a serious assessment looks beyond device placement into the operation (around 2:05). Later, I discuss why commercial facilities need a different plan from a home: more people, roles, and possible consequences. Those examples remain useful questions for a site walk today.

Schools need a usable plan

In this school-security conversation, I discuss assessments and the practical relationship between access control, cameras, and the way a campus operates. There is no universal device list. A useful plan reflects entrances, staff roles, visitor flow, and emergency procedures.

Media conversations also include NBC coverage of CTA camera deployment and WJLA's report on facial recognition and public safety. In each setting, the question was how the technology would actually help people investigate and respond.

Additional recordings include a technical discussion of surveillance frame rates and an ABC segment on regulated-facility security. Each addresses a different part of the field, from image performance to the way security must fit a facility's obligations.

Consulting when the decision is bigger than the equipment

A school, municipality, manufacturer, or multi-site business may have a room full of products and still lack a coherent security operation. I can help leadership examine the assumptions in a plan: what risks matter, where the current system fails, how staff would respond, what the organization truly owns, and which tradeoffs have been hidden by a glossy proposal.

My background spans software development, camera manufacturing, open VMS integration, public transit deployment, access control, alarms, emergency communication, and hands-on security system work. That breadth makes it easier to see where a decision made by one team becomes a burden for another.

What a useful engagement looks like

We first define the decision you need to make, the people affected, your existing environment, and what evidence is available. Depending on the assignment, the work can include a site or system review, interviews with operators, a vendor comparison, a technical briefing for leadership, and a prioritized set of decisions. The scope should fit the problem; I will not imply a full audit when a focused review is enough.

Typical contacts include superintendents, school operations leaders, municipal administrators, IT directors, facility managers, security directors, and executives responsible for a high-stakes technology investment. If you are trying to choose a vendor, recover from a poor installation, or understand whether an AI or cloud platform solves your problem, tell me the decision and its deadline.

Building panoramic video that operators could use

I started Sentry360 because conventional camera coverage left blind spots, while the early 360-degree systems often made operators work too hard to see a usable view. High resolution was only part of the answer. A panoramic image needed software that could dewarp it into familiar views, make the footage searchable, and connect it to the video management system a customer already used.

In a 2013 Security Today profile, I discussed building the company from a modest start and pushing the camera platform toward practical use. That history matters to me because it showed how much of a product's value lives beyond its spec sheet: integration, operator training, support, and the ability to review an incident later.

At the time, the security industry was moving from analog and fixed-view cameras toward megapixel IP video. A panoramic camera could capture the whole scene from a single position, then let an investigator look back and navigate the recorded view later. That retrospective PTZ capability was useful in a lobby or rail car where an event did not wait for someone to point a camera in the right direction. It also raised practical questions about pixel density, storage, lighting, and whether fewer cameras truly met the customer's evidence needs.

Making 360-degree video usable

Dewarping turns the curved view from a panoramic lens into perspectives an operator can navigate. The important questions are practical: can a person follow movement, see a face where necessary, and export the right evidence without learning a separate, closed workflow? Our dewarping SDK was designed to help connect panoramic video to established software. Industry coverage of the Sentry360 SDK documents that integration work.

I personally worked on the fisheye dewarping software and integration approach. Building that software changed how I talk to a customer about open architecture. The test is whether the camera, video management software, and operator workflow can work together under real conditions. Ask for a demonstration in the platform you intend to own, with sample footage from a scene like yours.

What made the SDK technically different

Contemporary industry reporting from 2009 described the SDK as a way for third-party systems to work with Sentry360 panoramic video across high- and low-level APIs. The report discussed C/C++, ActiveX, .NET, Windows, Linux, and raw-image flexibility. Those details matter because open architecture was a software and integration problem, not just a phrase on a product box.

One integration illustration on this page shows dewarped and panoramic views inside a video management environment. To me, that is the actual engineering test: can a customer's operators use the wide view in the platform they already rely on, live and after an incident?

Where the technology met public infrastructure

Real projects expose the distance between a technology demonstration and something staff can rely on. Sentry360's panoramic cameras were used in Chicago Transit Authority rail cars, where coverage, storage, maintenance, and evidence review have to work in a moving public environment. Read the published CTA deployment coverage; watch NBC's transit report.

Industry awards brought welcome attention to the engineering, but my most useful feedback came from seeing how an operator used the system. Recognition is a signal of technical ambition; it does not replace a site's own performance test.

Another published application involved panoramic cameras for Microsoft retail locations. The industry report on that deployment describes the technology and setting. That manufacturer-side experience informs how I evaluate current systems for Umbrella customers.

Early Sentry360 opportunities also included W Hotels, Des Plaines Public Works, and the Chicago Housing Authority, as described in the Security Today founder profile. The environments varied, but the same question came up repeatedly: who would use the video and what decision would it help them make?

From manufacturer to systems advisor

My focus now is the customer's whole security environment: cameras, access control, intrusion detection, emergency communication, software, network requirements, and the people who operate them. Umbrella designs, installs, and supports projects for schools, municipalities, and businesses in Chicago and Northern Illinois. We start by walking the site and asking what staff must see, control, and retrieve on an ordinary day and during an incident. Then we compare equipment against those tasks and the organization's ability to maintain it.

My advisory work benefits from seeing the same problem from the developer, manufacturer, integrator, and customer sides. The CTA deployment and retail applications are specific examples of work; they are not a substitute for examining a new client's setting.

For a school, that can mean testing how a door decision reaches the people who need it. For a municipal facility, it can mean making evidence retrieval and ownership clear. For a business, it can mean recognizing that a camera aimed at the wrong approach creates a record without an answer. See current Umbrella case studies for project details.

Research and views grounded in field evidence

Technology changes. The basic obligation does not: a system should help people understand what is happening and act in time. When I review a design, I ask where an incident would start, what evidence staff could see, what they could control, and who maintains the system after installation.

That last question is the subject of one of my short videos on cloud resilience. Watch the video directly, or use the Shorts carousel above.

AI should sharpen decisions, not bury people in alerts

Analytics for license plates, faces, weapons, or unusual activity need more than a compelling demo. Before using them in a school or public facility, I want to see how often they miss the event, how often they interrupt staff with a false alert, who verifies the output, and how the evidence is retained and accessed. A useful alert has an owner and a response path. Without those, automated detection can simply create a bigger queue of uncertainty.

For government and other public-sector systems, I also ask who can see the data, how long it is kept, how an export is logged, and whether a vendor can leave without taking the customer's records and workflows with it. Governance is part of the system design.

Field testing and honest recommendations

I want customers to hear the tradeoffs before they buy: what the equipment does well, what conditions make it underperform, what it costs to support, and where we would choose another approach. Umbrella's manufacturer reviews reflect field experience and product testing. They are a starting point for evaluation, not a promise that one manufacturer fits every site.

For advisory work, I start with the organization's risks, people, existing systems, and operating constraints. The useful answer may be a different design, a narrower upgrade, or a change in how a system is governed—not another product purchase.