Scaled Cognition Brief
B2B SaaS·Cx Agent Platform·Series A·🇺🇸🌐
Summary
Recently raised· Series A this monthScaled Cognition is a narrowing, enterprise‑grade play: an AI CX‑agent platform engineered for deterministic, policy‑compliant, hallucination‑free responses and sold into high‑assurance customer service use cases. It sits against broad incumbents like Sierra and Kore.ai and developer frameworks such as Rasa, but differentiates by selling predictability and compliance (SOC 2) — a wedge attractive to regulated or brand‑sensitive enterprises. Traction and credibility are strong: a $100M Series A led by Khosla Ventures (June 25, 2026), strategic commercial endorsement from Genesys (embedded Feb 10, 2026). The most important tension is commercial scaling: the company must convert technical defensibility into repeatable enterprise revenue at scale while incumbents and platform owners can replicate policy‑enforcement features rapidly.
Bull case
- $100M Series A led by Khosla Ventures on June 25, 2026 signals heavyweight investor conviction and provides a capital base to scale sales and engineering.
- Strategic partnership and product embedding with Genesys (announced Oct 16, 2025; embedded Feb 10, 2026) gives a credible GTM channel and enterprise integration proof‑point beyond standalone pilots.
- SOC 2 certification and a compact, research‑heavy founding team (Dan Roth, Dan Klein, Damon Pender) plus VP Research Anthony Platanios indicate both security posture and deep technical pedigree required for high‑assurance CX use cases.
Watch-outs
- Crowded competitive landscape: incumbents and specialists (Sierra, Kore.ai, Rasa, Decagon, Aisera) have broader product footprints or larger deployments and could out‑spend on GTM or replicate features.
- Platform risk: large platform owners and contact‑center vendors (including Genesys itself or Microsoft/AWS customer‑service stacks) can build native policy‑enforcement or hallucination‑mitigation features that undercut a standalone vendor’s value proposition.
- Commercial scaling tension after a large Series A: converting early proof‑points into repeatable enterprise ARR at the scale implied by a $100M raise is unproven and concentrates execution risk in sales and customer success.
- The company’s public materials do not state the regions or countries where it sells or operates, leaving geography, regulatory exposure, and go‑to‑market scope unclear.
From their website
Pricing
Not publicly disclosed. The company website does not list its pricing on any of the 6 pages we read.
Product suite
- APT-1
Agentic Pretrained Transformer model for building and running reliable, policy-compliant AI agents
- Agent Builder
Platform to create, test and deploy AI agents for text and voice interactions
- AgentTwin
- Vibe Builder (Studio • SDK)
- GenAPI™
- Agent Defender
- Agent Evaluator / Agentic Evaluation
- Agent Scorecards
- APT SDK
- VPC & On-Prem Hosting
Target segments
Compliance & certifications
Crunchbase
verified profile
- Scaled Cognition Raises $100M Series A Led by Khosla Ventures to Build Reliable Enterprise AIGlobeNewswire · 2026-06-25
- Scaled Cognition raises $100M to build AI that won’t hallucinateThe Next Web · 2026-06-25
- Scaled Cognition Raises $100M in Series A FundingFinSMEs · 2026-06-25
- Scaled Cognition Raises $100M Series A Led by Khosla Ventures to Build Reliable Enterprise AIVenture Capital Firms · 2026-06-25
- Scaled Cognition, a reliability-focused lab that develops the Agentic Pretrained Transformer model, raised a $100M Series A led by Khosla at a $750M valuation (Steven Rosenbush/Wall Street Journal)Techmeme · 2026-06-25
- Scaled Cognition Raises $100 Million to Address AI HallucinationsPYMNTS.com · 2026-06-25
Team
Scaled Cognition’s public roster names three co-founders (Dan Roth, Dan Klein, Damon Pender) with prior agentic-AI experience and a populated leadership and research-engineering team focused on building a hallucination-free CX agent platform.
The three co-founders (Dan Roth, Dan Klein, Damon Pender) have prior, explicit experience building agentic conversational systems (they co-founded Semantic Machines), giving direct domain fit for enterprise conversational AI.
The team combines a technical founding trio with extensive conversational-AI experience (Roth, Klein, Pender) and a full bench of research and engineering leaders; the main strength is deep domain research and production engineering. A material gap is no separately titled Head of Sales or General Counsel listed among leadership (the GTM bench is BD-focused rather than a named enterprise sales leader).
- Dan RothCo-Founder & CEODan RothCo-Founder & CEOSemantic MachinesTrinity College-HartfordCo-founded Semantic Machines (acquired by Microsoft)Repeat founder
- Dan KleinCo-Founder & CTODan KleinCo-Founder & CTOAdap.tv, Semantic Machines, Microsoft, University of California, BerkeleyStanford University, Cornell University (Bachelors, Computer Science, Linguistics, Mathematics), University of Oxford (Master's, Linguistics)Co-founded Semantic Machines (acquired by Microsoft)Repeat founder
- Damon PenderCo-Founder & CFODamon PenderCo-Founder & CFOSemantic MachinesCo-founded Semantic Machines (acquired by Microsoft)Repeat founder
Leadership
- Anthony PlataniosVP, ResearchAnthony PlataniosVP, ResearchCarnegie Mellon University
Leads the research organization responsible for the company’s reliability-focused model and evaluation tooling.
- Akshay KannanVP, ProductAkshay KannanVP, ProductUniversity of California, Berkeley
Oversees product strategy for the CX-focused, policy-enforced agent platform and associated tooling.
- Jordan DeLoachVP, EngineeringJordan DeLoachVP, EngineeringKansas State University
Leads engineering and technical operations for the platform; listed with a LinkedIn profile in public Crunchbase.
- Michael DelCielloChief Business Development OfficerMichael DelCielloChief Business Development Officer
Leads business development and strategic partnerships to place the company's reliability-focused agents into enterprise workflows.
- Matthew GardnerDirector, ResearchMatthew GardnerDirector, Research
Directs applied research efforts advancing model reliability and evaluation; appears in Crunchbase structured people as 'Research Scientist' with LinkedIn.
- Diana HoDirector, ProductDiana HoDirector, ProductInworld AI, Hello Sunshine, Microsoft, SAP ConcurMIT Sloan School of Management (MBA candidate; Gordon-Bluestein Fellowship, Entrepreneurship & Innovation), George Mason University (BS, Management Information Systems)
Manages product activities for the CX agent platform and tooling.
Ex top-tier - Evan BrownDirector, Business DevelopmentEvan BrownDirector, Business Development
Focuses on expanding commercial opportunities and partnerships for the platform.
- Jason BirchDirector, Business DevelopmentJason BirchDirector, Business Development
Works on business development alongside other GTM leaders to scale enterprise deals.
- Josh WalovitchDirector, MarketingJosh WalovitchDirector, Marketing
Leads marketing and positioning for the company's enterprise CX product and research story.
- Matt GardnerResearch ScientistMatt GardnerResearch Scientist
(Duplicate listing on Crunchbase and team page) Directs research and contributes to model reliability work.
Hiring concentrated in engineering and product (5 of 6 roles: Engineering 3, Product 2) with most hires as IC/mid (5 of 6) and only one senior, signaling a product-building stage focused on execution rather than scaling leadership or go-to-market.
Market Size
The closest published top-down figure for the addressable category (enterprise CX software/platforms) is Mordor Intelligence's global Customer Experience Management TAM at USD 22.79B in 2026. Narrowing that to enterprise-paid software (large-enterprise share × solutions share from the same report) yields an estimated serviceable addressable market of ≈USD 8.8B globally for enterprise CX platforms and solutions. Given Scaled Cognition's growth-stage profile (small team, SOC 2, and strategic investors), a plausible near-term obtainable scenario is ~0.5% of that SAM (≈USD 44M) over a ~3-year commercial expansion—this is a scenario, not a forecast.
- Mordor Intelligence ↗USD 22.79 Billion (Market Size in 2026) · Global · 10.31% CAGR
- TAMUSD 22.79B (Mordor Intelligence, 2026)Mordor Intelligence - Customer Experience Management Market
- Large-enterprise share (used to narrow to enterprise segment)60.76% of TAM (large enterprises, 2025 per Mordor Intelligence)Mordor Intelligence - Customer Experience Management Market
- Solutions share (to focus on software/platform revenue vs services)63.47% of TAM (solutions revenue share in 2025 per Mordor Intelligence)Mordor Intelligence - Customer Experience Management Market
- SAM formulaTAM × 60.76% (large enterprises) × 63.47% (solutions) = ≈ USD 8.79BMordor Intelligence segmentation and TAM
- SOM share0.5% of SAM over ~3 yearsGrounding: Scaled Cognition public facts — ≈34 employees; SOC 2; investors include Khosla and Genesys
- The published vendor figures for 'customer experience' vary by publisher and taxonomy; this sizing uses Mordor Intelligence's CEM definition and segment splits so SAM is internally consistent with that TAM. Other publishers (MarketsandMarkets, Grand View, Consainsights) report different universes and growth rates.
- No vendor-published figure specific to 'CX agent platforms' (agentic/AI conversational agent vendor revenue) was found; SAM is the enterprise-software slice of the broader CEM market rather than a narrow published 'agent platform' TAM.
- Scaled Cognition's public site discloses enterprise_custom pricing (no ACV); SOM therefore uses a share-of-SAM scenario anchored to public traction signals (employees, SOC 2, investors) rather than company-reported ARR or ACV.
Competitors
Top 5Niche, compliance-first CX agent with credible product differentiation but faces well-funded enterprise platform competitors.
Several well-funded platform incumbents (Sierra, Kore.ai, Decagon, Aisera) dominate enterprise CX; Scaled Cognition sits as a niche, compliance-focused challenger.
Market splits between massive enterprise platforms (Sierra, Kore.ai), developer frameworks (Rasa), and automation specialists (Decagon, Aisera). Scaled Cognition sits as a niche provider: enterprise CX agents focused on deterministic, policy‑enforced, hallucination‑free responses.
Scaled Cognition scaledcognition.com ↗ | Sierra sierra.ai ↗ | Decagon decagon.ai ↗ | Kore.ai kore.ai ↗ | Rasa rasa.com ↗ | Aisera aisera.com ↗ | |
|---|---|---|---|---|---|---|
| What they do | Building Super-Reliable Intelligence. | Enterprise AI CX agent platform | Agentic customer‑service platform (backend actions) | Conversational AI platform for enterprise bots | Open‑source conversational AI framework | AI customer & employee support automation |
| Segment | Your company | Direct | Direct | Indirect | Adjacent | Indirect |
| Stage | Series A | Series E | Secondary Market | Private Equity Round | Series C | Series D |
| Total raised | — | — | — | $275M | $71M | — |
| Valuation | — | $16B | $1.5B | — | — | — |
| Revenue | — | $200M | $35M | $100M–500M | $5M–25M | — |
| Lead investors | Khosla Ventures | Google Ventures, Tiger Global Management | Accel, Andreessen Horowitz, Bain Capital Ventures | AB Private Credit Investors | StepStone Group | Goldman Sachs, Thoma Bravo |
| Headcount | 34 | 488 | 469 | 1258 | 199 | 337 |
| Founded | 2022 | 2024 | 2023 | 2014 | 2016 | 2017 |
| CEO | — | — | — | Raj Koneru (Founder & CEO) | Melissa Gordon (CEO) | — |
| HQ | Amesbury, US | San Francisco, US | San Francisco, US | Orlando, US | San Francisco, US | Santa Clara, US |
| Geo focus | US | — | US, GB | US, IN, GB, JP +2 | US, DE | — |
| Ownership | Privately Held | — | Privately Held | Private | Private | — |
| Web traffic | 2K/mo ▲ | 183K/mo ▼ | 83K/mo ▼ | 266K/mo ▲ | 49K/mo ▼ | 15K/mo ▼ |
| LinkedIn followers | 7K | — | 65K | 106K | 25K | — |
| Traction | — | Broad enterprise adoption; major funding round | Targets 70–75% automation of support conversations | Recent private equity round | — | $131.9M reported annual revenue |
| Growth score | — | 91/100 | 70/100 | 89/100 | 73/100 | 52/100 |
| Heat score | — | 83/100 | 51/100 | 78/100 | 54/100 | 61/100 |
| Tech stack | 10 tools | 10 tools | 10 tools | 10 tools | 10 tools | 10 tools |
| Patents | 5 | — | — | — | — | 2 |
| Startup score | 62/100 | 62/100 | 62/100 | 62/100 | 62/100 | 50/100 |
| vs Scaled Cognition | — | Target wins on deterministic, policy‑enforced, hallucination‑free agents; Sierra is far ahead on scale, customer footprint, and funding. | Target offers stricter policy control and deterministic replies; Decagon is stronger on automating complex backend actions and scale. | Scaled Cognition offers stricter hallucination controls and compliance focus; Kore.ai is more established, broader in use cases and has recent private equity backing. | Scaled Cognition trades developer flexibility for an out‑of‑the‑box, policy‑enforced CX agent; Rasa is stronger for developer customization and self‑hosted control. | Scaled Cognition wins on policy enforcement and deterministic answers; Aisera is stronger on broad automation integrations and larger reported revenue. |
Investor overlap
- AccelDecagon, Rasa
- Andreessen HorowitzDecagon, Rasa
Recent funding in the space
- 2026SierraSeries E · $950M
- 2026Kore.aiPrivate Equity Round
Funding
- Seed Round$21 million2023Lerer Hippeau leadCrunchbase ↗
- Corporate RoundGenesys leadCrunchbase ↗
- Series A$100 millionJun 25, 2026Khosla Ventures leadGenesysGenesys Telecommunications Laboratories Inc.GlobeNewswire ↗
Public reporting shows Scaled Cognition completed a $100M Series A led by Khosla Ventures (June 25, 2026) with participation from Genesys, and a 2023 seed round (reported amount $21M) associated in Crunchbase with Lerer Hippeau. The itemized rounds sum to $121M.
- Crunchbase lists a Seed round with lead Lerer Hippeau but does not show an amount; press reporting (CO/AI) describes a $21M 2023 seed led by Khosla Ventures — the database and press disagree on the seed lead. The Seed entry uses Crunchbase as the canonical round-lead while the $21M figure is reported in press.
- The $100M Series A is reported in multiple press outlets and a GlobeNewswire company press release (June 25, 2026).
- Some aggregator/tracker pages referenced (Tracxn, Dealroom, PitchBook) list rounds or investor counts but provide no per-round primary notices; those aggregator claims were not used as primary evidence beyond the Crunchbase Crunchbase and the cited press release.
- Total ($121M) is the sum of itemized rounds; additional undisclosed or unreported rounds may exist.
Momentum
- FundingJune 25, 2026Scaled Cognition raised $100 million in a Series A round led by Khosla Ventures.GlobeNewswire ↗
- PartnershipOctober 16, 2025Genesys and Scaled Cognition announced a strategic partnership to advance reliable agentic AI for customer experience orchestration.Genesys ↗
- PartnershipFebruary 10, 2026Genesys unveiled an agentic virtual agent in its cloud platform that uses Scaled Cognition’s APT-1 model.Business Wire ↗
- Funding2023Scaled Cognition closed a $21 million Seed round in 2023.Crunchbase ↗
Scaled Cognition has recent high-impact signals: a $100M Series A led by Khosla Ventures (June 25, 2026) and a strategic partnership with Genesys (announced Oct 16, 2025) that has already been embedded into a Genesys product (Feb 10, 2026). Earlier seed funding (2023) shows prior investor traction.
- Coverage is concentrated in company and partner press releases and mainstream press reporting those releases.
- Some secondary outlets repackage the same GlobeNewswire/press-release content.
Customer Sentiment
7- Low latency / strong performancePartner Baseten reports 40% lower overall latency and <120 ms time-to-first-token after integrating Scaled Cognition's stack, highlighting execution and infrastructure performance as a market differentiator.“40% lower overall latency. <120 ms time to first token.”Baseten case study ↗
- Deterministic, hallucination-free behavior / policy adherenceThe company emphasizes models and platform features designed to eliminate hallucinations and enforce company policies and deterministic constraints for CX workflows.“Scaled Cognition offers a model that ensures customer interactions are based on verified answers that adhere to company policies.”Crunchbase ↗
- High benchmark / leader-board performanceReported leaderboard and benchmark wins for APT-1 on agentic/complex task benchmarks, presented as evidence of technical leadership in agentic workflows.“APT-1 topped agentic leaderboards, outperforming existing frontier models from OpenAI, Anthropic, and Google.”Baseten case study ↗
- Fast development / rapid deploymentAgent Builder platform is presented as enabling developers to create, test, and deploy agents quickly (claimed under an hour) and to validate agents without connecting to real APIs.“companies can build, test, and deploy specialized AI agents within an hour—without integrating with real APIs during development.”CO/AI (getcoai) ↗
- Synthetic training data pipelineThe platform uses a fully synthetic agentic data pipeline to generate grounded training data without human labeling, framed as reducing implementation risk and data requirements.“They've built a data pipeline that generates precisely the grounded data needed for agent training.”CO/AI (getcoai) ↗
- Integration & deployment flexibilityPlatform and models support hybrid cloud, VPC/on-prem deployment options and integration with existing support platforms, positioned for enterprise security and compliance needs.“Hosted or VPC/On-Premise deployment options”AI Agents Directory ↗
- Strong investor / partner signalProminent investors and partners are cited (e.g., Khosla Ventures lead investor; Baseten partnership; Genesys partnership) supporting go-to-market traction and technical credibility.“Khosla Ventures led their $21 million seed round in 2023”getcoai / Baseten case study / Crunchbase ↗
Public third‑party material (press articles, a Baseten case study, directory and company profiles) presents Scaled Cognition positively: technical claims focus on deterministic, hallucination‑free agent behavior, strong benchmark performance for APT‑1, low-latency execution, fast agent development/deployment, and enterprise deployment flexibility. Customer/partner quotes (Baseten) praise latency improvements and collaboration.
- No independent consumer review platform scores or large-volume user reviews found; available sources are press coverage, a partner case study, directory listings, and company profiles.
- Most customer-facing evidence is partner testimonial and company/press claims rather than broad user reviews.
AI Visibility
We ask ChatGPT WITH web search about the company and show its answers plus the sources it cited — the real AI-search experience. The key signal is share of voice: does the company come up when AI is asked about its space, not just when asked by name.
Models asked ChatGPT · web search
Accuracy of what AI said: Accurate — Correctly identifies Scaled Cognition's focus on reliable, deterministic CX AI, founders/leadership, products (APT-1, GenAPI), Berkeley base, 2022 founding, and Series A backing.
Scaled Cognition is a company specializing in developing reliable and verifiable artificial intelligence (AI) solutions tailored for enterprise-scale operations. Their primary focus is on creating AI systems that deliver consistent and accurate results, addressing the challenges of reliability and determinism in AI applications. (scaledcognition.com)
Founders and Leadership: The company was co-founded by Dan Roth (CEO), Dan Klein (CTO), and Damon Pender (CFO). Their leadership team includes individuals with extensive experience in AI research and development. (scaledcognition.com)
Products and Services: Scaled Cognition offers APT-1, a frontier model designed for customer service applications. APT-1 is engineered to eliminate hallucinations and ensure policy compliance, providing reliable AI-driven customer interactions. The company also provides tools like GenAPI™ for simulating and validating AI agents before deployment, ensuring they perform correctly in real-world scenarios. (scaledcognition.com)
Company Size and Stage: Founded in 2022 and based in Berkeley, United States, Scaled Cognition has been operational for approximately four years. The company has secured Series A funding led by Khosla Ventures, indicating a growth phase with a focus on expanding research and team development.
Notable Facts:
- Scaled Cognition emphasizes "Super-Reliable Intelligence," aiming to provide AI solutions that are deterministic and free from hallucinations, which is crucial for enterprise applications where consistency and accuracy are paramount. (scaledcognition.com)
- The company is backed by leading investors, including Vinod Khosla and Genesys, reflecting strong support and confidence in their mission to advance reliable AI technologies. (scaledcognition.com)
For more detailed information, you can visit their official website:
Here is a concise list of notable companies and products in the customer experience (CX) agent platform space:
- Adobe Experience Platform (AEP): Offers AI-powered tools like Audience Agent and Journey Agent to enhance customer experiences and marketing effectiveness. (techradar.com)
- Salesforce Agentforce Contact Center: Integrates CRM, AI agent technology, and telephony to automate and streamline customer support. (itpro.com)
- RingCentral RingCX: Unifies customer conversations across phone, chat, SMS, and digital channels into a single environment. (ringcentral.com)
- Nextiva: Provides a digital contact center platform with voice, SMS, chat, video, and CRM integration, featuring AI-enhanced call intelligence and sentiment scoring. (nextiva.com)
- Zendesk Support: A cloud-based customer service solution consolidating interactions from multiple channels into a unified platform. (techradar.com)
- Tidio: An all-in-one customer service platform offering live chat, AI-powered chatbots, and multichannel messaging for small and medium-sized businesses. (techradar.com)
- Cresta: Unifies automation, agent assist, and conversation intelligence on one foundation to enhance customer experience. (cresta.com)
- Certainly: An agentic AI platform built for customer experience, offering multi-model flexibility and omnichannel deployment. (certainly.io)
- Quiq: Provides an enterprise customer-experience platform using agentic AI, integrating with CRM systems to support digital customer interactions. (en.wikipedia.org)
- Minerva CQ: A real-time Agent Assist product deployed in voice-based customer support, integrating real-time transcription, intent and sentiment detection, and contextual retrieval. (arxiv.org)
These platforms are recognized for their contributions to enhancing customer experience through innovative agent technologies.
In the customer experience (CX) agent platform space, several companies offer innovative solutions to enhance customer interactions through AI-driven technologies. Here are some notable platforms:
Adobe CX Enterprise Introduced in 2026, Adobe's CX Enterprise is an AI-powered platform designed to manage the full customer lifecycle—from acquisition to loyalty—in a unified system. It integrates AI agents, reusable workflows, and real-time customer data via the Adobe Experience Platform. (techradar.com)
Salesforce CEO criticizes Microsoft's AI tools, touts new Agentforce latform Salesforce Agentforce Launched in 2025, Agentforce is deeply integrated into Salesforce's Service Cloud, allowing agents to access CRM data, case history, knowledge base, and product catalog natively. It's particularly beneficial for enterprises already within the Salesforce ecosystem. (certainly.io)
Genesys Cloud CX、24年度第1四半期に年間経常収益10億ドルを突破 | Genesys Cloud Services, Inc.のプレスリリース Genesys Cloud CX Tailored for large contact centers managing high inbound voice volume, Genesys Cloud CX offers robust voice capabilities and comprehensive customer experience solutions. (certainly.io)
NiCE CXone | CX Integration, Management & Support from SVL NICE CXone Mpower NICE's CXone Mpower provides customer service automation and has been recognized for its AI-driven orchestration solutions, enhancing customer service tasks through automation. (en.wikipedia.org)
Decagon Announces Commercial Pilot with Deutsche Telekom and Strategic Investment from T.Capital to Fuel Enterprise Growth - Silicon Canals Decagon Decagon is an AI agent platform that uses Agent Operating Procedures (AOPs) to handle interactions across chat, email, voice, and SMS. It's particularly strong for tech and financial-services companies running Zendesk or Salesforce as the helpdesk layer. (atomicwork.com)
These platforms represent a range of solutions in the CX agent platform space, each offering unique features tailored to different business needs.
AI finds Scaled Cognition when asked by name and describes its mission and products accurately, but does not surface the company unprompted among notable CX agent platforms. The company's description from the AI can be trusted for basic facts but it is not yet prominent in unprompted category recommendations.
- AI answers can vary run-to-run; presence in category or recommendation lists may change over time.
- Web visibility for young B2B AI companies can be limited, so unprompted omission does not imply lack of traction.
Risks & Red Flags
No material public risk signals found. An absence of public concerns isn’t a clean bill of health — early or private companies may simply not generate coverage.
Absence of public reports is not a clean bill of health for a private/early-stage company; limited public visibility or recent founding can produce little third‑party coverage even when operational risks exist.