InfoComm 2026 was in Las Vegas. We were there. The same themes kept showing up across the floor: complete solutions, AI doing the work inside hardware, installs getting cleaner. And one demo handed us a piece of fruit and made us reach for it. Here is what mattered.
The demo handed us an apple. We reached for it. The apple was not real. Take a second with that.
Google Beam is a turnkey immersive experience: microphones, speakers, cameras, and display in one packaged solution. No custom build required. The clearest winner of mind-blowing technology on the floor.
Executives have been asking for this for years. The industry kept answering with custom projects and budget conversations that ended the meeting. Google Beam is a defined experience. An integrator specifies it and installs it. Dense computing and subscription costs are part of the conversation. So is the apple. Brief clients accordingly.
Runs natively in Google Meet and Zoom. No proprietary codec. The multi-camera array and lenticular display produce a light-field effect: eye contact, spatial depth, things flat-panel video does not do. Dense compute is required at the infrastructure level. Factor that into the deployment conversation early. New category. Not an iteration on anything that existed before. Integration patterns will develop fast now that hardware is shipping.
The Intelligent Video Bar is built on Intel/Linux hardware with integrated network switching inside the unit. It is not a video bar. It is a compute platform that also handles your meeting room.
Most AV rooms run on a collection of individual devices. A codec, a camera, a DSP, a control processor, a network switch. Something is always the weak link. The Shure Intelligent Video Bar consolidates conferencing, audio processing, camera control, control panel and network switching into one unit. Fewer things to manage. Fewer vendors to call. That math is not complicated.
Intel/Linux architecture. Network switching and DSP built into the chassis. The design allows for the possibility of a future secondary camera, advanced room workflow triggers, and scheduling and occupancy integration. Shure is building toward a room OS. The hardware gives them the foundation to do it.
Pixelhue and Pixelhue Pro AV handles AV-over-IP routing, signal processing, KVM, control, and LED wall management. In one platform. We have been ahead of this technology as a team. The rest of the market is catching up. That gap is still real.
Specifying a command center, a broadcast-lite all-hands space, or any room with multiple camera sources and an LED wall? The Pixelhue platform removes a meaningful amount of rack gear from the equation. Fewer integration points. Simpler commissioning. A cleaner conversation with the client about what they are buying and why.
4x 12G-SDI inputs. 8x DisplayPort/HDMI outputs. 4x primary HDMI outputs. Native Dante. The U5 and U5 Mini extend the platform into smaller deployments without changing the architecture underneath. Router, processor, and wall controller in one managed unit. That is the whole story.
The 1 Beyond i12D. Two 4K PTZ cameras working together to track whoever is speaking. No operator. No manual camera control. The AI lives on the box. Not in the cloud. Not on a server somewhere down the hall.
Medium and large conference rooms have a camera problem. One wide-angle lens captures everyone and flatters nobody. The i12D fixes that. A reference camera and a dedicated 4K PTZ follow the active speaker automatically. 12x optical zoom. The room does not have to do anything
Dual 4K PTZ camera system with a reference camera and integrated microphone array for spatial audio detection. AI processing runs entirely on-device. No external compute required, no cloud dependency, no latency introduced by offloading to a remote platform. 12x optical zoom. Designed for medium to large spaces where a wide-angle lens alone does not provide the close-up coverage the room needs. Speaker tracking triggers from the microphone array, not from motion detection, which means it follows the right person, not the one who just stood up to grab a coffee.
Dual 4K PTZ system. Reference camera. Integrated microphone array for spatial audio detection. AI runs on-device. No cloud dependency, no external compute, no latency. Speaker tracking triggers from the microphone array, not motion detection. It follows the right person, not the one who just stood up to grab a coffee.
Sony showed up with a full hand. BRAVIA BZ-P Series displays, AI PTZ cameras, Face Recognition, AI Content Generation. All demonstrated together as one coherent story.
The BRAVIA BZ-P Series runs 43 to 85 inches across three tiers. Centered VESA. Uniform bezels. Deep black anti-glare. The AI PTZ cameras track presenters and frame for sports without an operator. Face Recognition adds tablet-based presenter ID with intelligent camera follow. The AI Content Generation tool for digital signage was working on the floor, not a prototype. The examples were early, but it was live and functional. For clients who want one vendor covering displays and cameras without the ecosystem falling apart, Sony is the answer right now.
AI processing runs on-device. Not in the cloud. PTZ tracking and sports framing work regardless of network conditions. The Face Recognition system links presenter identity to camera position without a human in the loop. The AI Content Generation platform points at a device-level AI that could reduce dependence on third-party signage platforms over time. Early. Worth watching.
The NetSpeek AI Service Agent identifies AV and network issues, diagnoses the cause, and remediates automatically. No ticket required. An AI focused on AV that actually understands room and network configuration. The category winner for Day 2 monitoring.
Most AV monitoring tells you something broke. NetSpeek tells you why, then fixes it. Layer it with a service ticketing platform and you have a managed services offering that monitors and diagnoses an entire organization's AV infrastructure. Clients will hesitate at anything labeled AI on their network. That hesitation is shrinking. The platform is early. The direction is not.
Auto-remediation separates this from monitoring-only tools. It corrects network configuration issues and AV device states. It does not just flag them. The architecture points toward ticketing and service desk integration. Pairing NetSpeek with a structured service delivery workflow is where the real managed services story gets built. The question is not whether this category wins. The question is who gets there first.
Two Sennheiser products made the Top 10. The 48-Channel Bidirectional Wireless Platform is the one that addresses environments where channel count is the design constraint.
Large venues, higher education lecture halls, enterprise conference centers. The ones that have historically required stacking multiple wireless systems to hit their channel targets. One platform, 48 bidirectional channels, simpler RF coordination, smaller rack footprint. The scalability path does not require adding another system on top of what is already there.
Bidirectional means IEM and microphone channels share the same infrastructure. Relevant for houses of worship and performance venues where sends and returns both matter. RF coordination across 48 channels in a congested environment is where the engineering story holds or does not. Real installs will answer that faster than any spec sheet.
The TCC M Ceiling Microphone. Circular and square versions. Aggressive price point. Performance, aesthetics, and price in the same product. That combination has historically required picking two.
Ceiling microphones get value-engineered out of mid-tier conference rooms. Not because they are not better. Because they cost more. The TCC M changes that. If you have been specifying boundary microphones or goosenecks because the ceiling mic option did not pencil out, revisit the spec.
Circular and square form factors both available. Matters for retrofit projects where the ceiling tile grid is not negotiable. Performance specs compete against solutions that cost significantly more. If it holds up in real-world installs, the adoption curve for ceiling microphones in enterprise environments accelerates.
Neat's MCP Integration puts a natural language interface on top of Neat endpoints. You talk to the devices instead of navigating a portal. The portals were not popular.
'Is the camera in Conference Room B online?' Plain English in. Plain English out. No portal login. No navigating to the right device page. No interpreting a status field. For IT teams managing distributed rooms without dedicated AV staff, this is a different way to work
MCP, Model Context Protocol, is an emerging standard for giving AI systems structured access to external tools and data. Neat's implementation exposes device state, configuration, and diagnostic data through a natural language layer. Future platforms built on this will not require trained operators to navigate proprietary interfaces. Early. Clear direction.
The Rally AI Camera and Rally AI Camera Pro. Single-wire connectivity. Flexible coverage. New in-wall mounting option. The Pro adds a dual-camera system with 15x hybrid zoom for larger rooms.
One cable. In practice it holds. For clients where the room has to look right, law firms, executive suites, high-end corporate spaces, a single wire disappearing into the wall is the difference between an install that looks like AV gear and one that looks like it was always there. Reliable tracking. Less time on-site
Single-wire over USB-C or ethernet depending on configuration. The in-wall mounting option opens cleaner paths in new construction and renovation. The Pro's dual-camera system with 15x hybrid zoom covers rooms where a wide-angle lens alone does not do the job. AI tracking runs on-device. No external processing.
The Top 10 got the rankings. The floor had more. These did not make the cut. Some of them probably deserved to.
Inogeni U-BRIDGE USB-C and U-BRIDGE USB-C DUO: Among the first USB-C extenders to support DisplayPort Alt Mode. Full USB-C: video, USB 2.0, power, and ethernet up to 70 meters over a single CAT6A cable. The DUO adds a second HDMI output for dual-display setups. BYOM rooms need one cable at the table that handles everything. This one does.
Novastar COEX Sending Units and Receiving Cards: 10G networking. Larger panels at the same pixel pitch, or denser solutions at the same panel size. Overhauled software across the lineup. An open, publicly documented API that simplifies multi-source wall configurations.
Planar Cobra and Dragon LED Wall: 0.6mm pixel pitch. EverPixel dual-LED technology: two LEDs per color. One fails, the paired LED compensates at full brightness. Dead pixels disappear. Off-axis color retention is exceptional. Possibly the most visually impressive display on the floor. Reliability, price point, visual performance. Hard to argue with all three at once.
ModusVR: Tools we were not fully aware of until the show. Strides being made toward dragging in floor plans and BOMs directly. New features on the horizon worth a closer look with the team.
RDL D-NVC2 Dante Volume Control: Single-gang wall plate. Rotary encoder. Color LCD. Controls one or two Dante channels. No DSP. No programming. Someone in a breakout room just needs a volume knob. This is the volume knob.