Personal AI

What Personal AI does

Personal AI (legal entity Human AI Labs, Inc.) is a memory-driven AI platform company founded in 2020 and headquartered in San Francisco. The company's core product is a proprietary Memory Core architecture that orchestrates five primitives — Encoding, Stabilizing, Storing, Retrieving, and Updating — across episodic, semantic, working, relationship, temporal, and procedural memory types. Built on top of Memory Core are Persona Studio (a no-code builder for AI personas) and My AI (an end-user assistant), all powered by Personal AI's Memory-Based Small Language Models (SLMs), branded as Personal Language Models (PLMs). The platform is positioned as a privacy-first, memory-native alternative to general-purpose LLMs, validated by NVIDIA Research's 'Small Language Models are the Future of Agentic AI' paper.

The technology is deployed across cloud, regional compute hubs, and edge sites via NVIDIA's AI Grid framework, with claimed benchmarks of 15ms Time to First Token (27-67x faster than cloud LLMs), $0.02 per million output tokens (40x cheaper than Gemma-27B), and sub-500ms end-to-end voice latency. Personal AI sells primarily to telecommunications carriers (Comcast, AT&T, Verizon, T-Mobile, Singtel) embedding SLMs directly into carrier networks and edge hardware, with secondary segments in Fortune 500 enterprises (Amer Sports/Salomon, BNI, WWT), mid-market law firms (Three Crowns, Flatiron Law with Stanford CodeX), SMBs (via carrier-delivered AI Receptionists), and edge devices (Snapdragon X Series NPU PCs).

The company monetizes through enterprise SaaS subscriptions on multi-year contracts, usage-based carrier token billing (claimed 92% gross margin), Google Cloud Marketplace listings, and custom Persona professional services. Consumer Personal Plan subscriptions were paused in January 2025 for a strategic enhancement period. As of December 2025, the company closed a Series A led by Differential Ventures, and recently announced a flagship strategic partnership with Comcast and NVIDIA at GTC 2026 to deploy its memory-based SLMs on the AI Grid.

Name

Personal AI

Legal name

Human AI Labs, Inc.

Website

https://personal.ai

Company type

Private

Founded year

2020

Operating status

Operating

Headcount range

51–100 employees

Short description

Personal AI is a memory-driven AI platform that provides carrier-native Small Language Models and infrastructure for telecom operators, Fortune 500 enterprises, and developers, enabling persistent AI identity with sub-500ms latency and 40x cheaper token economics across cloud and edge deployments.

Ownership category

akta.pro rank

Product category

AI Memory Platform

NAICS

Computer Systems Design and Related Services (54151), Custom Computer Programming Services (541511), Software Publishers (5132)

SIC

Services-Prepackaged Software (7372), Services-Computer Integrated Systems Design (7373)

akta.pro primary industry

End-to-End Enterprise AI Platforms (MLOps & Model Lifecycle Management) (HDAEANAA)

akta.pro secondary industries

AI Application Enablement Platforms (Copilot/Agent Frameworks, SDKs) (HDAEANAJ), On-Device Inference Runtimes & SDKs (mobile/embedded) (HDAAAJAB)

Keywords

• Memory-based AI platform
• Small language models
• Edge AI infrastructure
• Distributed inference platform
• AI persona software

Headquarters

HQ city

San Francisco

HQ country

United States

HQ region

North America

Offices4 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Marketing or Sales, Infrastructure, Operations

Revenue model

1. Enterprise SaaS Subscriptions: Platform Master Subscription Agreement (SaaS use rights) with enterprise customers on multi-year subscription terms. Customers pay Fees in US dollars on each Order Form, non-refundable, with personal AI retaining all rights to the underlying Platform IP. Custom-built Personas developed per customer use case.
2. Carrier Token / Usage-Based Billing: For telecom carriers, Personal AI offers Tokens billed to the line — the 'fourth primitive' alongside Talk, Text and Data — metered at claimed 92% gross margin. Distributed inference on the AI Grid achieves $0.02 per million tokens versus $0.80 for Gemma-27B, enabling telecom-grade token economics.
3. Cloud Marketplace Distribution (Google Cloud Marketplace): Personal AI Workforce Platform is listed on the Google Cloud Marketplace, enabling enterprise self-service procurement as an additional SaaS revenue channel.
4. Professional Services & Custom Persona Development: Personal AI develops custom Personas for enterprise customers as specified on each Order Form, with non-exclusive Display licenses for the duration of the Subscription Term. Customers wishing to commercialize Personas must execute Personal AI's standard reseller agreement, opening a potential licensing/reseller revenue stream.
5. Consumer Personal Plans (Per-Seat): Personal AI offered Personal Plan subscriptions to individual consumers (paused as of January 23, 2025 for a strategic enhancement period), indicating a per-seat consumer subscription revenue stream.

Pricing tiers

Go-to-market motion6 records

Distribution channels6 records

Marketing channels11 records

Core offering

Personal AI builds a memory-driven AI platform for telecom carriers, Fortune 500 enterprises, and developers, anchored by Memory Core (unified context and memory service), Personal Language Models (PLMs/SLMs), and Persona Studio (no-code persona builder). The platform replaces stateless LLM context retrieval with persistent, identity-bonded memory across episodic, semantic, working, relationship, temporal, and procedural memory types. It is deployed via cloud, regional, and telco edge infrastructure on the NVIDIA AI Grid with sub-500ms latency and 40x lower token cost than general-purpose LLMs.

Product overview

Personal AI is a single unified platform-plus-modules architecture organized in three layers, from infrastructure to end-user product, all powered by persistent evolving memory. The foundation layer is **Memory Core**, the unified context and memory service that encodes, stabilizes, stores, retrieves, and updates persistent memory for self-improving AI; the builder layer is **Persona Studio**, a no-code platform to design, configure, and deploy AI Personas without writing code; and the application layer is **My AI**, an end-user assistant that knows the user, remembers every conversation, learns preferences, and evolves over time. Together, Memory Core, Persona Studio, and My AI compose the Memory Platform that Personal AI delivers to telecom operators, enterprises, and developers, and that powers use cases such as the AI Receptionist, connected phone/robot/vehicle experiences, and integrations with NVIDIA's AI Grid, HPE ProLiant, Lanner Edge AI servers, and Snapdragon X Series NPUs.

Differentiator

Problem solved

Functional benefit

Brands

• My AI: Application-layer product offering a long-running personal AI assistant that remembers every conversation and evolves with the user over time.
• Persona Studio
• Memory Core

Products and services

• My AI

Quantifiable outcome

• 27x faster Time to First Token (15ms vs. 1,000ms cloud LLM baseline) on the Comcast AI Grid
• +8 more outcomes

Named customers13 records

Segments6 records

Ideal customer profiles6 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration13 records

AI capability15 records

Feature9 records

Partnerships

Twelve partnerships are on record, tiered flagship, core and minor.

• NVIDIAflagshipStrategic or Co-development Partner · 18 March 2026NVIDIA and Personal AI are co-developing the AI Grid reference architecture for distributed edge inference. Personal AI's Memory-Based Small Language Models are validated on NVIDIA RTX PRO 6000 GPUs in the Comcast AI Grid trials (announced at NVIDIA GTC 2026, March 17-18, 2026). The collaboration spans GTC keynote delivery, benchmark co-publication, and joint telecom-grade AI monetization strategy. NVIDIA Senior Executive Ronnie Vasishta delivered a keynote on AI and telecom infrastructure at SF Tech Week (Oct 2025).
• Hewlett Packard Enterprise (HPE)flagshipTechnology or Integration · 18 March 2026HPE launched the HPE AI Grid, a new infrastructure platform aligned with NVIDIA's AI Grid reference architecture, designed to securely connect AI factories and distributed inference clusters across regional and edge sites. HPE's ProLiant servers (combined with Juniper telco-grade networking and NVIDIA GPUs/DPUs) are the deployment substrate for Personal AI's SLMs in the Comcast trial.
• ComcastflagshipStrategic or Co-development Partner · 17 March 2026Comcast announced a first-of-its-kind collaboration with NVIDIA and Personal AI to test AI workloads running at the edge of its nationwide distributed network reaching 65 million homes and businesses. Three initial use cases include personalized advertising (Decart), a small business concierge agent powered by Personal AI's SLM on HPE ProLiant servers, and ultra-low-latency gaming streaming. Initial lab testing 'demonstrated strong performance' validating latency, power, cost, resiliency, and scalability improvements.
• DecartcoreTechnology or Integration · 17 March 2026Decart is a co-trial participant in the Comcast/NVIDIA/Personal AI edge AI field trial. Decart provides real-time AI video models powering a personalized advertising agent use case alongside Personal AI's small business concierge agent and NVIDIA's GeForce Now low-latency gaming.
• Lanner ElectronicsflagshipOEM/ Whitelabel/ Licensing Partner · 28 October 2025Lanner Electronics and Personal AI announced a strategic partnership at NVIDIA GTC DC to deliver a combined hardware and software platform enabling telecom providers to deploy advanced AI services at the network edge. The integrated solution combines Lanner's NVIDIA MGX-certified Edge AI servers with Personal AI's Small Language Model technology, enabling telcos to offer AI services at 90% gross margins with 20x better cost efficiency than traditional LLMs. Purpose-built for AI Radio Access Networks (a building block for 6G infrastructure), targeting over 350 million telecom subscribers in the US.
• NVIDIA RivacoreTechnology or Integration · 29 April 2025Personal AI integrates NVIDIA Riva (speech AI SDK) to power AI Receptionists for small business owners, enabling voice-based conversational AI on NVIDIA's speech platform.
• Stanford University CodeX CenterminorStrategic or Co-development Partner · 21 March 2025Stanford University's CodeX center partnered with Personal AI and law firms Three Crowns LLP and Flatiron Law Group LLP to develop platforms like DealMentor and DealDriver that simulate senior attorneys' reasoning for training associates, providing feedback, and analyzing contracts.
• Google CloudcoreChannel Partner/ Reseller/ Distributor · 26 September 2024Personal AI brought its AI Workforce Platform to the Google Cloud Marketplace on September 26, 2024, enabling enterprise self-service procurement and consumption of Personal AI's platform against cloud commitments.
• Former Congressman and OpenAI Board MemberminorOthers · 10 July 2024A former US Congressman and OpenAI Board Member joined Personal AI's Board of Directors in July 2024 (announced via Personal AI press). Specific individual not named in the source material.
• MicrosoftcoreTechnology or IntegrationMicrosoft is featured as a Personal AI partner and customer logo. The Microsoft Copilot vs Personal AI comparison (March 25, 2025) frames Personal AI's persona-driven, single-subscription, memory-native approach vs. Microsoft's general-assistant integration within the Microsoft 365 ecosystem.
• Amazon Web Services (AWS)coreTechnology or IntegrationAWS is featured as a Personal AI partner and customer logo. Personal AI is positioned as a Cloud Hyperscaler partner per the Partner Network page.
• WWT (World Wide Technology)coreImplementation/ SI/ Consulting PartnerWWT is a featured Personal AI partner and customer logo — a major global technology solutions provider that delivers enterprise-scale AI infrastructure and implementation services.

Scale indicators14 records

Recent moves7 records

Expansion highlights9 records

Direct peers

• Sierra AI: Sierra builds conversational AI agents for enterprise customer experiences with persistent memory and persona capabilities, directly comparable to Personal AI's Memory Core and Persona Studio offerings. Both target Fortune 500 enterprise buyers with AI persona technology and emphasize brand-aligned, memory-grounded responses over generic LLM chat.
• Cresta: Cresta provides real-time generative AI for contact centers and customer service, with persistent agent memory and AI-receptionist-style capabilities that overlap directly with Personal AI's AI Receptionist product. Both companies pursue enterprise and SMB voice-AI deployments with sub-second latency requirements.
• Letta (formerly MemGPT): Letta is building memory-augmented LLM agent infrastructure that closely parallels Personal AI's Memory Core architecture, focused on giving AI agents persistent long-term memory. Both companies compete for the memory-layer position in the agentic AI stack and sell to AI engineers and developers building persona-driven applications.

Broad incumbents

• Moveworks: Moveworks provides enterprise AI assistants for IT, HR, and employee support at Fortune 500 scale, comparable to Personal AI's enterprise persona deployments. While Moveworks focuses on employee-facing workflows, both companies sell memory-grounded AI agents into large enterprise customers with SOC 2/HIPAA-grade compliance postures.
• Cohere: Cohere is an enterprise-focused foundation model company with private deployment options, comparable to Personal AI's positioning as a private, programmable alternative to frontier LLMs. While Cohere offers general-purpose LLMs and Personal AI focuses on memory-based SLMs, both target regulated enterprise buyers with data sovereignty and customization requirements.
• AI21 Labs: AI21 Labs offers enterprise AI platforms including task-specific SLMs, comparable to Personal AI's Personal Language Models positioning as a specialized alternative to general LLMs. Both companies pursue enterprise and developer channels for domain-specific AI deployments rather than consumer chat.

Emerging players

• Inworld AI: Inworld AI builds character-based AI with persistent long-term memory for gaming, entertainment, and brand experiences, directly comparable to Personal AI's memory-grounded persona technology. Both companies invest heavily in identity-bonded memory architectures and developer tooling for persona creation.
• Cognee: Cognee provides an open-source memory layer for AI agents that orchestrates episodic, semantic, and procedural memory types - architecturally analogous to Personal AI's Memory Core five-primitives model. Both companies sell into the AI engineer and developer community building memory-augmented agentic applications.
• Hume AI: Hume AI builds emotionally intelligent voice AI with sub-second latency and personality customization, comparable to Personal AI's voice SLM and AI Receptionist capabilities. Both companies target enterprise and SMB voice AI deployments with custom persona technology and measured latency advantages over OpenAI Realtime.
• Edge Impulse: Edge Impulse is an edge AI development platform for deploying optimized models on edge devices, comparable to Personal AI's positioning of memory-based SLMs on edge hardware (HPE ProLiant, Lanner MGX servers, Snapdragon NPUs). Both companies serve the same emerging category of edge-deployed AI for telecom, IoT, and embedded use cases.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat7 records

Key risks7 records

Key highlights7 records

Customer concentration

• Xtwitter.com/personalai_
• YouTubeyoutube.com/channel/UC_h69pwGH3iF3RVPpBag1bw
• Facebookfacebook.com/personalai
• LinkedInlinkedin.com/company/personalai
• Instagraminstagram.com/personalai

Compliance4 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles10 records

Funding overview

Funding rounds10 records

Investors10 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Personal AI