Coram AI Bet $35 Million That Buyers Won't Replace Their Cameras

On 11 June 2026, Coram AI raised a $35 million Series B led by Ansa Capital and Battery Ventures, taking its total funding to $66 million. The round came after Coram quadrupled revenue and tripled its customer count since a $13.8 million Series A the year before, according to the company's own funding announcement.

Coram doesn't sell cameras. It sells software that reads footage from cameras a buyer already owns. The Series B is evidence that buyers will pay specifically for that, and not for new hardware.

The Numbers Behind The Round

Coram says it now serves more than 1,500 sites across the US and Canada, a footprint it credits to tripling its customer base and quadrupling revenue since the Series A, according to the company's funding post. Named customers of Coram's platform include Hershey's Ice Cream and Lakepointe Church, a Dallas-area network with more than 30,000 congregants across eight campuses, alongside Fortune 500 companies, school districts and healthcare providers the company declines to name individually.

"Most security systems just record what happened," chief executive Ashesh Jain said of the pitch behind that growth. "Only later, after a manual search, might you find the incident." That line is the whole argument: Coram is selling the search, not the camera that captured the footage.

The company says the new capital goes mostly toward sales capacity, plus continued work on the AI product itself and on customer support, rather than toward building or subsidizing hardware. That spending pattern matches the pitch: a software company scaling a software sale.

Who Coram Actually Serves

Coram AI's core claim is that it works with any IP camera already installed, no rip-and-replace required. The company also says it integrates with roughly 99 percent of existing access control systems, per its own site.

That claim matters most to one kind of buyer: someone who already has a camera investment and doesn't want to write it off. Schools, churches, healthcare clinics and warehouse operators fit that description, and the product reflects it.

The platform pairs natural-language video search with what it calls Deep Investigation, autonomous agents that compile a report across multiple cameras instead of making a person scrub hours of footage. Real-time alerts cover firearms, falls, PPE violations and unauthorized entry, and the company's own site lists SOC 2 Type II and HIPAA certification, relevant to the healthcare and school buyers it names as customers.

The company also claims a site can go live within minutes of signup, since there is no new camera to mount, wire or configure on the network. That speed claim is the flip side of the compatibility claim below: both rest on the cameras already being there.

The Camera Vendors Have The Most To Lose

Verkada, the most visible name in AI-enabled security cameras, built its business on selling its own hardware, cameras, access control panels, intercoms and sensors, alongside a cloud management platform, according to Verkada's own site. That model works as long as buyers accept that better AI means new hardware.

Coram's pitch removes that assumption. If buyers believe it, the vendors whose margin depends on the camera sale lose the argument, not a feature race. The same logic hits the guard company billing for a person watching sixteen monitors: Coram's autonomous investigation is built specifically to replace that manual review step.

Coram Isn't Ambient.ai, Overview AI, Or GeoEvident

AI physical security search results lump together products that solve different problems. Ambient.ai is the closest peer on paper, computer vision layered onto existing enterprise cameras. But it is built for large corporate security operations centers, not the school district or church campus Coram is chasing.

Overview AI is not a security platform at all. It puts AI on the production line to catch manufacturing defects, and picking it over Coram for perimeter security misreads what it does.

GeoEvident works in the opposite direction: identifying where an existing photo was taken, for investigation and verification, rather than monitoring a live feed. All three meet Coram only at the edges. Treating any of them as interchangeable is the mistake a rushed vendor comparison makes.

The Pitch Has One Untested Claim

The weakest point in Coram's story is the part nobody outside the company has verified: the 99 percent compatibility figure and the any-camera promise. Coram doesn't publish which camera brands or firmware versions the claim excludes.

A buyer running an older or less common recorder system won't know whether they're in the 99 percent until a technician shows up. Pricing is quote-only and licensed per camera, so the total cost of keeping your hardware isn't comparable to a hardware vendor's quote without a sales call on both sides.

If that number lands close to a hardware-replacing system over a multi-year term, the argument that started this round of funding gets a lot less interesting. For a buyer trying to weigh that tradeoff against other options without booking a string of demos, that's the kind of constraint-matching question Smart Match exists to help answer.

What To Watch By The End Of 2026

Coram has not published per-camera pricing, and that is the one number that would let a buyer stack it directly against a hardware vendor's number, no demo required. Watch whether that changes before the end of 2026.

A company confident that dropping the hardware requirement is discount enough to win on its own has reason to publish the number. One that still needs a live conversation to close the gap does not.

Until then, the strongest evidence for Coram's pitch isn't on its website. It's the more than 1,500 sites that signed on without ever seeing a public price list.

Frequently asked questions

What does Coram AI actually do?

Coram AI adds an AI layer, natural-language video search, autonomous "Deep Investigation" reports across multiple cameras, and real-time alerts for threats like firearms or unauthorized entry, on top of IP cameras a business already has installed. It does not require replacing existing cameras, and the company says it integrates with roughly 99 percent of existing access control systems.

How much does Coram AI cost?

Coram AI's pricing is quote-only and licensed per camera; the company does not publish a public price list. Buyers need to request a demo to get a quote, which makes it hard to compare directly against a hardware vendor's list price without going through both sales processes.

Is Coram AI different from Verkada?

Yes. Verkada sells its own proprietary cameras, access control panels and sensors alongside its own software platform, so upgrading typically means buying new hardware. Coram is software-only and pitches itself as working with cameras a buyer already owns, which is the core distinction between the two.

Who is Coram AI built for?

Coram's named customers, including Hershey's Ice Cream and Lakepointe Church, point to mid-size organizations like schools, churches, healthcare clinics and warehouse operators that already have camera infrastructure. Buyers wanting a single enterprise-scale security operations center vendor for large corporate campuses may be better served by Ambient.ai.

How much funding has Coram AI raised?

Coram AI has raised $66 million total, including a $35 million Series B announced on 11 June 2026 and led by Ansa Capital and Battery Ventures. The company said the round followed 4x revenue growth and a tripled customer base since its $13.8 million Series A a year earlier.

Covered in this guide

• Coram AI: AI-native physical security platform: plain-English video search, gun detection and access control layered onto any IP camera. 4.9/5 on G2 from 189 reviews. Per-camera license, quote only.
• Ambient.ai: Ambient.ai is a computer vision platform that flags 150+ physical security threats via existing enterprise cameras rather than requiring new hardware.
• GeoEvident: GeoEvident is an AI geolocation intelligence platform that identifies locations from photos using computer vision. Designed for OSINT, law enforcement, and investigations. Cloud or on-premise deployment.
• Overview AI: Overview AI builds edge vision cameras with 150+ TOPS of on-device NVIDIA GPU compute to catch manufacturing defects on the production line.

Sources