Amplifier Security

• Amplifier could remain without an arm's-length named reference, since every customer named in the reviewed record dates to April 2024, four of the six also appear in its own founder, angel or advisor network, and the practitioner videos published since name no account as a customer.
• Jamf could ship the same employee conversation itself, because Amplifier both names it as an alliance partner and routes findings and fixes through it.
• Amplifier's employee-routed workflows could stall on employee response, since those fixes depend on people answering an agent in Slack or Teams and the 94% engagement rate behind that claim is Amplifier's own.
• The absence of a published SOC 2 or ISO attestation could stall enterprise procurement, since the platform reads identity, endpoint and vulnerability data across the workforce.
• Buyers could keep struggling to place Amplifier in a budget line, since agentic Workforce Security, Autonomous User Security and self-healing all appear as its label across its own pages.

Amplifier Security sells to the team that owns a finding it cannot close alone. Its own statement of the gap is blunt: every security team can spot the gaps, and few can finish the fix when it requires action from employees. The four buyers it addresses by name are the CISO who owns the risk number, the security engineer who owns the fix, the IT team that owns the queue, and the awareness leader who owns training.

Independent coverage frames the same gap, and largely in the founders' words. A Forbes contributor piece quotes Sadalgi saying that cybersecurity has a last-mile problem, that organisations have bought the tools and employees do not always use them properly. Intellyx's analyst brief describes the condition from the buyer's side, as a plethora of tools each generating alerts that are difficult to correlate to what an individual person actually did.

What neither supplies is a measurement. Every figure sizing the gap in the reviewed record sits on Amplifier's own pages, which makes the problem statement credible and the scale of it vendor-asserted. [s1, s8, s9]

The platform's unit of work is a person rather than an asset. Its Human Risk Graph is described as a per-employee risk profile that ingests signals from existing identity, security and productivity tools and maps devices, applications, identities and behaviours, which is what lets an agent address one employee about one finding instead of issuing generic guidance.

Two named agents and a builder run on that profile. Ampy Security Engineer reaches the employee in Slack, Microsoft Teams or the browser to explain what is wrong, why it matters and how to fix it. Ampy Risk Advisor surfaces findings for security and IT leaders, explains the rationale, recommends next steps, and acts only once a leader approves. AI Agent Studio is a drag-and-drop builder where a team sets which findings trigger an agent, what employees see, how reminders escalate and which tools carry out the remediation.

The connectors are what make the profile possible, and the platform page names 25+ of them across Jamf, Okta, CrowdStrike, Microsoft Intune, Tenable and Automox. VentureBeat described the same architecture in 2024 under different names, as a security data fabric, a security hub and an engagement studio. The reviewed record documents the design and carries no third-party technical evaluation and no customer technical account of how the agents behave in production, so a buyer evaluating the claim has the vendor's description and a demo to work from. [s5, s3, s7]

Amplifier names its own competitive set. A comparison post it published in January 2026 lists Living Security, Hoxhunt, Proofpoint ZenGuide, Mimecast, SoSafe and CybSafe, and argues that each covers one slice of employee risk while Amplifier covers devices, identity and behaviour together. The post is the company's own, so it establishes who Amplifier sells against rather than how the products compare. VentureBeat reported separately in April 2024 that Amplifier planned to take on Zscaler-owned Avalor Security and Silk Security.

The harder contest is with the vendors it reads from. Amplifier's architecture claim is that it sits between existing tools and the workforce, not inside any silo, and the tools it lists include Jamf, Okta, CrowdStrike, Microsoft Intune, Tenable and Automox. Jamf is also the alliance partner named in Amplifier's seed announcement. Each of them either produces a finding Amplifier consumes or applies a fix Amplifier requests, and any of them could add an employee conversation to a product the customer already pays for.

The argument Amplifier makes against that is span. Its own framing is that vulnerability management, endpoint detection, identity and awareness each cover a slice and none of them reaches the human who has to act. The reviewed record states that position and does not test it against any rival's coverage, so a buyer weighing it has to check the span claim against their own stack. [s10, s1, s3, s4]

The named customers in the reviewed record come from one week in April 2024. VentureBeat reported that Amplifier had worked with security teams at more than 15 companies to design the product and that seven of those partners had converted into early customers, of which it named six: Guardant Health, Skyflow, Instabase, Oscar Health, BetterCloud and Zemoso. A Forbes contributor piece the same day described about 15 marquee design partners ranging from Instabase to TikTok, and quoted Guardant Health's head of security.

Four of those six names also sit in Amplifier's own network. VentureBeat records that both founders came from BetterCloud, which is on the customer list; Skyflow's co-founder Anshu Sharma appears on Amplifier's angel list; and the CISOs of Guardant Health and Oscar Health appear on its advisor list. A design-partner base overlapping the founders' network is ordinary at pre-seed, and it means these names do not by themselves show the product winning buyers who came to it cold.

What followed names practitioners rather than accounts. The April 2025 seed announcement pairs an alliance partnership with Jamf with a quote from the CISO of Automox, a company Amplifier lists as a connector, and the March 2026 launch reports averages across unnamed customers: policy acknowledgment moving from 75% to 98.5%, EDR and DLP coverage from 60% to 98%, and 7,500 endpoint vulnerabilities remediated in under 30 days. Amplifier published every one of those figures, and the public record does not say which customers they came from. [s7, s8, s2, s4, s5]

Both founders came out of the same company. VentureBeat identifies Shreyas Sadalgi and Thomas Donnelly as former BetterCloud executives, and Amplifier's about page lists them as co-founder and chief executive and co-founder and chief technology officer. Sadalgi's route there ran through security: Crunchbase records him as director of business development at Blue Coat from 2008 and senior vice president of business development at Centrify from 2012, and as an advisor to Banyan Security. It also records a single founded organization for him, Amplifier itself, and the reviewed record credits neither founder with a prior security product of their own.

The names around the company are better known than the names inside it. Amplifier's angel list names Arctic Wolf co-founder Brian NeSmith, Slack co-founder Cal Henderson, OneTrust co-founder Kabir Barday and Skyflow co-founder Anshu Sharma. Its advisor list names a former Amazon CISO and the sitting CISOs of Guardant Health and Oscar Health. TechOperators led the seed round, and its general partner Daniel Ingevaldson is named as a director on Amplifier's SEC filing.

For a buyer that roster is a signal about access rather than about the product. It overlaps the design-partner list and it sits alongside a company that reached security leaders early, and the reviewed record connects none of it to how the platform is built or run. [s17, s7, s2, s4, s15]

No SOC 2, ISO certification or other attestation appears anywhere in the reviewed sources or on the surfaces probed for one. Amplifier's terms of service list audit information such as SOC2 and penetration-testing reports among the categories the parties must keep confidential, which says what happens to such a report if one is shared and not that one exists. The reviewed record also names no regime that requires either for this product class.

The controls Amplifier does describe are its own. It states that security and IT leaders review and approve actions before execution, that AI Agent Studio governs what can trigger an agent and where human approval is required, and that the system works only from the structured context of the risk graph and within remediation paths the security team defines. All of that is the vendor describing its own design, and the reviewed record carries no outside review of it.

The product's position creates its own question for a buyer. To build a per-employee profile it ingests identity, endpoint, vulnerability, SIEM, SaaS-security and training signals, and to close a finding it messages employees inside the company's own Slack or Teams, or in their browser. That is a cross-tool view of the workforce plus a channel into it, extended to a company incorporated in 2023, with no third-party assurance over either in the reviewed record. [s12, s16, s5, s4, s15]

Add analyzed competitors to compare them side by side with Amplifier Security.

Amplifier segments by who owns the unclosed finding, not by industry. Its site sells four separate views to four buyers inside the same account: the CISO who has to show the board that risk fell, the security engineer who owns the fix, the IT team whose queue absorbs it, and the awareness leader who owns training. That is one product sold four ways rather than four products.

The size band is stated plainly in the company's own writing. Its January 2026 comparison post is addressed to mid-market teams and names their specific bind, enterprise-grade risk carried on startup-grade headcount. That framing fits the product, because the argument for paying someone to reach employees is strongest where nobody has the staff to chase them.

The reviewed record describes none of the 2024 design partners as a mid-market buyer. It names Guardant Health, Oscar Health and, in a Forbes contributor piece, TikTok, which that piece groups among consumer-facing businesses alongside technology specialists such as Instabase. Nothing in the record gives the size of any of them. It also does not say which segment buys today, so the stated target and the named logos are not the same evidence.

The capability that carries the product is attribution, not conversation. Before an agent can ask anyone to patch anything, Amplifier has to know whose laptop it is, and the platform page sells exactly that: map every device to an owner, resolve ownership conflicts before they become gaps, no orphaned endpoints. The Human Risk Graph then hangs identity, endpoint, vulnerability, SIEM, SaaS-security and training signals off that person.

AI does two jobs on top of it. Ampy Security Engineer holds the exchange with the employee in Slack, Microsoft Teams or the browser, explaining what is wrong and why it matters before walking them through the fix. Ampy Risk Advisor does the reverse for leaders, surfacing findings in a digest, explaining the rationale and recommending actions it executes only after approval.

Amplifier is explicit that the AI is bounded rather than autonomous. It states that the system works only from the structured context of the risk graph, acts within remediation paths the security team defines, and that AI Agent Studio controls what can trigger an agent, what it may do, and where a human must approve. The company's own comparison post calls the result human-in-the-loop self-healing, where employees fix their own issues with AI-guided help rather than the software acting silently.

A cross-customer learning asset is plausible here and is not in the record. A platform that watches which message, channel and escalation actually closes which finding across many customers could learn what a single tenant cannot, and the reviewed sources describe no such loop and name no data Amplifier keeps beyond each customer's own signals.

The motion runs through practitioners the founders could reach, and Amplifier described it that way at the time. Its chief executive told VentureBeat the strategy was to build a brand with modern security practitioners at iconic companies who could articulate the pain and the differentiation, and that this would drive peer-to-peer effects. VentureBeat reported seven conversions and named six of them, and no source since labels any account a customer.

Partnership is the other channel the record shows. The April 2025 seed announcement pairs the round with an alliance partnership with Jamf, and the quote it carries alongside the investors is from the CISO of Automox, which Amplifier lists as a connector rather than presents as a customer. The record describes joint conference presence, at Jamf's RSA Conference booth, and no sales motion for the Jamf partnership.

What the site does not carry is a self-serve path. The reviewed record carries no published price and no product trial. Amplifier's own Resources menu lists interactive demos and an ROI calculator, and its home page offers a five-minute self-assessment that needs no infrastructure access, so a buyer can size the problem before a call but not the spend. For a mid-market buyer that means the evaluation starts with a call rather than a signup.

Amplifier publishes no price. The reviewed sources carry no pricing page, no packaging tiers and no unit of charge, and the site's call to action throughout is to book a demo. A buyer cannot size the spend before contacting sales.

The value argument is stated in savings rather than in price. The home page claims an average annual saving of $1.2 million from tool consolidation, a figure Amplifier publishes and the reviewed record does not corroborate. That claim is the shape of a business case built for a buyer who has to justify a new line item, and the record does not show what produced the number.

The bundled engineering complicates the comparison. Amplifier states that its service includes forward-deployed engineers who help implement agents and supervise edge cases, so the contract carries people as well as software, and the record does not separate what either costs.

Delivery starts with connectors and the vendor states it is quick. The platform page names 25+ native connectors across Jamf, Okta, CrowdStrike, Microsoft Intune, Tenable and Automox, and the integrations page adds custom webhook triggers for the rest. Amplifier's own comparison post concedes that the product gets the most value once connected to the existing stack, while claiming deployment runs in days rather than months.

Operating it is a two-sided job. On the security side a team builds agents in a drag-and-drop studio, setting triggers, employee-facing messages, escalation steps and which tool performs the remediation. On the employee side the product is a conversation in Slack, Microsoft Teams or the browser, which means adoption depends on people answering a bot about their own laptop.

Amplifier does not leave that to the customer alone. It states that the service includes forward-deployed security engineers who implement agents, supervise edge cases and tune performance, which says both that the agents need tending and that Amplifier expects to do the tending.

No SOC 2, ISO certification or other attestation appears anywhere in the reviewed sources or on the surfaces probed for one. Amplifier's terms list audit information such as SOC2 and penetration-testing reports among the categories the parties keep confidential, which says what happens to such a report if one is shared rather than that one exists. No regime requiring either for this product class appears in the reviewed record.

What the company does publish is a description of its own guardrails. Security and IT leaders review and approve actions before execution, the agent studio controls what may trigger an agent and where approval is required, and the system draws only on approved enterprise signals inside bounded remediation paths. Every one of those statements is Amplifier's, and the reviewed record contains no outside review of any of them.

The trust question is sharpened by what the product needs. To profile an employee it ingests identity, endpoint, vulnerability, SIEM, SaaS-security and training data, and to close a finding it speaks to that employee inside the company's own Slack or Teams. A buyer is granting a company incorporated in 2023 both a cross-tool view of its workforce and a channel into it, with no third-party assurance over either in the reviewed record.

Amplifier's platform strategy is to be the layer above other platforms. Its own architecture claim is that it sits between existing tools and the workforce rather than inside any silo, and its integration list reads as a map of the stack it depends on: Jamf, Okta, CrowdStrike, Microsoft Intune, Tenable, Automox, Slack and Microsoft Teams.

That position is a dependency as much as a strategy. The findings it acts on and the tools that carry out the remediation belong to other vendors, so Amplifier's coverage is bounded by what those vendors expose. Its own delivery paths are broader than one channel, because Ampy reaches employees in Slack, Microsoft Teams and the browser. The company argues the offsetting benefit is span, that vulnerability management, endpoint detection, identity and awareness each cover a slice and none reaches the human who has to act.

Ecosystem placement so far runs through one partner. Jamf is the one alliance partner the reviewed record names, which puts the ecosystem signal in a single vendor rather than across the stack Amplifier integrates with.

The founding pair is one team from one prior company. VentureBeat identifies Shreyas Sadalgi and Thomas Donnelly as former BetterCloud executives, and Amplifier's about page lists them as chief executive and chief technology officer. Crunchbase records Sadalgi's earlier roles as business development at Blue Coat and Centrify, plus a single founded organization for him, Amplifier itself. The reviewed record adds nothing about Donnelly's track record beyond that BetterCloud role.

The company staffs against its own delivery risk. Its about page names a head of engineering and an application security architect among the staff, and the March 2026 launch commits forward-deployed security engineers to customer accounts. The about page names a head of engineering and an application security architect alongside the founders, for a company that carries both a platform and a services obligation.

The investors and advisors bring the track record the founders' own histories do not. Arctic Wolf, Slack, OneTrust and Skyflow co-founders sit on the angel list, a former Amazon CISO and the sitting CISOs of Guardant Health and Oscar Health sit on the advisor list, and TechOperators general partner Daniel Ingevaldson is named as a director on the SEC filing. That bench explains the access this company has had, and the reviewed record ties none of it to how the product is built.