What Gradient Labs does

Gradient Labs is an AI-native customer operations platform built exclusively for regulated financial services. Founded in 2023 in London by three former Monzo AI and data leaders (CEO Dimitri Masin, CTO Neal Lathia, Chief Scientist Danai Antoniou), the company deploys a suite of specialist vertical agents — Lending, Disputes, Voice, Outbound and KYB — that work together in a multi-agent architecture to automate frontline support, proactive outreach and back-office workflows across banks, lenders, insurers, fintechs and remittance providers. The company reports 20+ active customers in production (including Wise, Monzo, a large European digital bank, Current, Stash, Rho, Zego, Plum, Pockit, SteadyPay, Yonder, Nala, Penfold, Flagstone, Morse/Sling Money, Bondora, nsave and Rain), serving 32M+ end users across the UK, Europe and the US following its October 2025 US market entry via a New York office.

The underlying platform is a procedural-agent framework that orchestrates frontier LLMs from Anthropic (Claude 3.5 Sonnet and later), OpenAI (GPT-4.1, GPT-5.4 mini/nano per OpenAI's own customer story) and Google, with multi-LLM failover across cloud providers. Every conversational turn runs 20+ finance-specific guardrails covering FCA Consumer Duty, CONC, Reg E/Z, PSD2, GDPR, EU AI Act, FDCPA, TCPA, Reg F and related rules; the platform ingests text, audio (30+ languages) and structured account data, with auto-extracted "Facts" mined from historical support conversations and a human-in-the-loop "Ask a Human" handoff for edge cases. Integrations span helpdesks (Intercom, Zendesk, Salesforce, Freshdesk), contact centres (Twilio, RingCentral, Genesys, Nice), lending systems (LoanPro), card-scheme dispute rails (Visa Resolve, Mastercard, Amex) and core banking processors (Fiserv, FIS). The company holds SOC 2 Type 2 certification and is GDPR-ready.

Gradient Labs sells on an outcomes-based model with no platform fees and a money-back deployment guarantee, layered with usage-linked charges on high-volume Lending and Disputes workloads. GTM is primarily enterprise direct sales to regulated financial institutions, with a light self-serve entry point on existing helpdesks, heavy event presence at fintech conferences, a bank-led innovation channel via the NatWest 2026 FinTech Programme, and an explicit US field expansion motion run out of New York. The company has raised a £2.8M seed (Aug 2024, LocalGlobe), a $13M Series A (Jul 2025, Redpoint Ventures) and a $26M Series A extension (Jun 2026, Octopus Ventures and CommerzVentures), for cumulative disclosed funding of $42.6M and an estimated ~$60M valuation at the 2025 round.

Name

Gradient Labs

Legal name

Gradient Labs Limited

Website

https://gradient-labs.ai

Company type

Private

Founded year

2023

Operating status

Operating

Headcount range

51–100 employees

Short description

Gradient Labs is an AI-native customer operations platform for regulated financial services, deploying specialist AI agents for lending, disputes, voice, outbound and KYB workflows. The London-headquartered company serves 20+ production customers including Wise, Zego, Plum and SteadyPay across UK, Europe and the US.

Ownership category

akta.pro rank

Product category

Conversational AI for Financial Services Customer Operations

NAICS

Custom Computer Programming Services (541511)

SIC

Services-Prepackaged Software (7372)

akta.pro primary industry

Financial Services Personalization (Banking, Investing, Insurance) (HDAAAGAH)

akta.pro secondary industries

Enterprise AI Governance, Risk & Compliance Platforms (Model Risk, Audit, Policies) (HDAEANAE), Model Serving, Inference & Deployment Platforms (APIs, Edge/On-Prem) (HDAEANAC)

Keywords

• Conversational AI platform
• AI customer operations
• Fintech automation software
• Lending lifecycle automation
• Financial services compliance AI

Headquarters

HQ city

London

HQ country

United Kingdom

HQ region

Europe

Offices2 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

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

Revenue model

1. Outcomes-based AI agent subscription: B2B SaaS revenue from financial institutions, charged on a simple outcomes-based model with no platform fees — customers pay only for successful query resolutions that the AI agent delivered. Tiered against deployment scope and resolution volume.
2. Performance-guarantee / money-back deployment model: Once a use case is scoped, Gradient Labs guarantees the deployment outcome — if it fails to deliver, the customer gets their money back, aligning revenue with realized performance and de-risking enterprise adoption.
3. Embedded lending and back-office usage fees: Usage-linked charges for higher-volume Lending and Disputes workloads (e.g. outbound collections calls, chargebacks processed) where volume is the primary commercial axis, layered on top of the outcomes-based subscription.

Pricing tiers

Go-to-market motion5 records

Distribution channels5 records

Marketing channels8 records

Core offering

Gradient Labs sells an AI-native customer operations platform for regulated financial services, delivered as a suite of specialist AI agents (Lending, Disputes, Voice, Outbound, KYB) that run on a procedural-agent framework with finance-specific guardrails applied on every turn. Customers access the platform via a web app or public API, integrating it into existing helpdesks, contact centres, lending management systems and card-scheme rails to automate frontline support, proactive outreach and complex back-office workflows such as chargebacks, hardship assessment, KYB, and collections.

Product overview

Gradient Labs provides an AI-native customer operations platform for financial services, organized as a platform-plus-specialist-agents architecture. The core platform, branded Otto (the Gradient Labs AI Agent), is configurable through a web app or via the Gradient Labs public API and acts as the operating system that runs between the specialist agents. On top of this base, Gradient Labs sells a suite of specialist agents — the Lending Agent, the Disputes Agent, the Voice AI Agent, the Outbound Agent, and the KYB Agent — each purpose-built for one regulated job and working together as a multi-agent system across frontline support, proactive outreach, and back-office operations. The Lending Agent anchors the lending lifecycle (SteadyPay, Pockit, LHV); the Disputes Agent handles chargeback workflows (Yonder, Wise); the Voice AI Agent powers phone support; and the Outbound Agent drives collections, fraud alerts, and verification. Supporting features include the Ask a Human handoff, Procedures for multi-step account actions, and an outcomes-based pricing model. The platform is consumed by 20+ active customers in production across retail banking, remittances, lending, insurance, savings, and pensions, serving 32 million end users with integrations into Intercom, Zendesk, Salesforce, Freshdesk, Twilio, RingCentral, Genesys, Nice, LoanPro, Visa, Mastercard, Amex, Enfuce, Thredd, PayPal, Klarna, Fiserv, and FIS.

Differentiator

Problem solved

Functional benefit

Brands

• Otto: Gradient Labs' flagship AI agent brand, used across deployments as the named conversational AI persona handling customer operations.

Products and services

• Otto (Gradient Labs AI Agent Platform) The flagship AI customer operations platform for regulated financial services, branded as the 'Otto' agent, configurable via the Gradient Labs web app or public API. Operates as the multi-agent system under which specialist agents (Lending, Disputes, Voice, Outbound, KYB) work, handling frontline support across voice, chat, email and SMS plus back-office investigations and proactive outreach with finance-specific guardrails applied on every turn.
• Lending Agent Specialist agent that automates the entire borrower lifecycle end to end for lenders and embedded credit platforms: applications, welcome and onboarding, active-borrower support, hardship and forbearance assessment, inbound and outbound collections, and back-office work. Handles 100,000+ voice calls per month across lending customers and is pre-configured for FDCPA/TCPA/Reg F (US), FCA Consumer Duty/CONC/Breathing Space (UK) and GDPR/EU AI Act (EU), with native integration into LoanPro lending management systems.
• Disputes Agent Specialist agent that automates the full chargeback and dispute lifecycle for card issuers, payment platforms and fintechs: intake via voice/text/email, classification against card-scheme reason codes, evidence review, follow-up to fill gaps, decisioning with human sign-off, and direct chargeback submission to Mastercard scheme APIs and Visa Resolve Online. Reports 98% accuracy, 24/7 coverage, a decrease in average resolution time and £30+ saved per case, with 20+ dispute-specific guardrails applied on every turn.
• Voice AI Agent Voice agent purpose-built for regulated financial services phone workflows. Handles complex calls with finance-specific guardrails (false promises, tipping-off, vulnerable-customer handling) on every turn, supports 30+ languages and regional accents, integrates into back-office systems to resolve cases end-to-end over the phone, and can be provisioned without requiring a dedicated phone line. Used for collections, hardship, fraud alerts, active borrower support and 24/7 customer service.
• Outbound Agent Proactive outreach agent that contacts customers on voice, SMS and email to handle collections, document collection, fraud prevention, verification and other agent-initiated tasks in financial services. Runs the same finance-specific guardrails as inbound interactions, honours callback requests and can leave voicemails. Generally available from February 2026.
• KYB Agent Back-office specialist agent that runs identity and document checks for business verification (KYB). Checks each document against internal policy, requests anything missing from the customer, screens for sanctions, and verifies the business before routing it to onboarding. Referenced as one of the suite of specialist agents in the company's Series A funding announcement.

Quantifiable outcome

• 80–90% peak resolution rate in mature deployments; 98% CSAT outperforming human teams
• +9 more outcomes

Named customers20 records

Segments6 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration18 records

AI capability14 records

Feature9 records

Partnerships

24 partnerships are on record, tiered core technology partner, day-1 integration, standard integration, voice / messaging integration, contact centre integration, lending system integration, card scheme integration, card processing integration, payments integration, core banking / processing integration, flagship uk bank partner, pilot banking partner and high-profile individual ambassador.

• Anthropiccore technology partnerTechnology or IntegrationFrontier LLM provider — Gradient Labs' agents are built on Anthropic's Claude 3.5 Sonnet and subsequent models; co-published customer story on Claude.com. Multi-LLM stack also uses OpenAI and Google models, but Anthropic is publicly highlighted.
• OpenAIcore technology partnerTechnology or IntegrationFrontier LLM provider powering parts of Gradient Labs' agent stack (e.g. GPT-4.1, GPT-5.4 mini and nano), per OpenAI's own customer story; featured as a partnership blog post in May 2026.
• Googlecore technology partnerTechnology or IntegrationMulti-model AI strategy includes Google's frontier LLMs alongside OpenAI and Anthropic, ensuring failover and access to the latest models.
• Intercomday-1 integrationTechnology or IntegrationCustomer-facing helpdesk connector — Gradient Labs can be deployed on Intercom with no engineering work on day one, and Plum's deployment is built on Intercom.
• Zendeskday-1 integrationTechnology or IntegrationHelpdesk connector enabling day-1 deployment of Gradient Labs' agents without engineering work, and integration for richer procedure automation.
• Freshdeskday-1 integrationTechnology or IntegrationHelpdesk connector enabling day-1 deployment of Gradient Labs' agents with no integration work required.
• Salesforcestandard integrationTechnology or IntegrationCRM/helpdesk integration allowing Gradient Labs to plug into existing customer support workflows.
• Twiliovoice / messaging integrationTechnology or IntegrationTelephony and messaging provider integrated into Gradient Labs' omnichannel agent stack.
• RingCentralcontact centre integrationTechnology or IntegrationContact centre integration that supports Gradient Labs' omnichannel voice agent deployments.
• Genesyscontact centre integrationTechnology or IntegrationContact centre integration enabling enterprise voice deployments of the AI agent.
• Nicecontact centre integrationTechnology or IntegrationContact centre integration enabling enterprise voice and customer-experience deployments.
• LoanProlending system integrationTechnology or IntegrationLending management system integration supporting the Lending Agent's borrower-lifecycle workflows.
• Visacard scheme integrationTechnology or IntegrationIntegration with Visa Resolve Online for chargeback submission and dispute handling by the Disputes Agent.
• Mastercardcard scheme integrationTechnology or IntegrationIntegration with Mastercard chargeback APIs for end-to-end dispute submission, including via Resolve Online.
• American Expresscard scheme integrationTechnology or IntegrationIntegration supporting Amex chargeback and dispute workflows handled by the Disputes Agent.
• Enfucecard processing integrationTechnology or IntegrationCard-processing and BIN-sponsor integration supporting European issuer deployments.
• Threddcard processing integrationTechnology or IntegrationCard-processing partner integration enabling issuer-side dispute and account workflows.
• PayPalpayments integrationTechnology or IntegrationPayments platform integration supporting dispute and account workflows for merchants and fintechs using PayPal rails.
• Klarnapayments integrationTechnology or IntegrationBNPL/payments integration supporting dispute and customer-operation workflows.
• Fiservcore banking / processing integrationTechnology or IntegrationCore banking / processing integration supporting enterprise customer operations deployments.
• FIScore banking / processing integrationTechnology or IntegrationCore banking / processing integration supporting enterprise customer operations deployments.
• NatWest Groupflagship uk bank partnerStrategic or Co-development PartnerSelected for NatWest's 2026 FinTech Programme (12-week AI-focused sprint themed "How AI is Shaping the Future of Customer Experience"); receives mentorship from senior leaders and access to NatWest's innovation ecosystem, with a path to live pilots inside the bank.
• LHV Bankpilot banking partnerStrategic or Co-development PartnerPilot partnership using Gradient Labs' agentic AI for email customer support to test efficiency and accountability ahead of broader rollout at LHV Bank.
• Tom Blomfield (individual)high-profile individual ambassadorOthersFormer Monzo CEO serves as Gradient Labs' "Ambassador to the UK Government" — provides public endorsement and policy access for the company. Not a commercial channel partner.

Scale indicators14 records

Recent moves7 records

Expansion highlights6 records

Direct peers

• Sierra AI: Sierra builds vertical AI agents for enterprise customer experience, with a similar emphasis on deep domain specialization, guardrails and autonomy. It is the closest direct competitor in philosophy and enterprise GTM motion to Gradient Labs.
• Decagon: Decagon provides an AI customer service platform for enterprises with a focus on reliability, customisation and integration into existing support stacks. It overlaps directly with Gradient Labs' multi-agent platform approach and serves regulated enterprise customers.
• Forethought: Forethought offers AI agents for customer support workflows (Triage, Assist, Resolve, Discover) and competes for the same enterprise support-automation budget. While more horizontal than Gradient Labs, its Resolve and Assist products target the same autonomous-agent use case.
• Kasisto: Kasisto's KAI platform is a financial-services-specific conversational AI used by banks for customer service and engagement. It is a direct vertical peer, albeit historically more focused on digital assistants than full back-office automation like Gradient Labs' Lending and Disputes Agents.
• Posh Technologies: Posh builds AI assistants specifically for banks and credit unions (Knowledge Assistant, Conversational Phone Assistant, Frontline Assistant). It targets the same banking customer base as Gradient Labs, with a comparable vertical-only GTM.
• PolyAI: PolyAI is a voice-first conversational AI platform for enterprise customer service, including banks and financial services. Its production-grade voice agents compete directly with Gradient Labs' Voice AI Agent for regulated phone workflows such as collections and customer support.
• Cognigy: Cognigy provides an enterprise conversational and agentic AI platform for customer service automation across voice and digital channels. It serves regulated industries including financial services and competes for the same enterprise agent platform budget.
• Cresta: Cresta offers real-time AI agents and coaching for contact centres, including financial services deployments. Its focus on voice/chat automation in regulated enterprises makes it a comparable peer to Gradient Labs' Voice and Outbound Agents.

Broad incumbents

• ASAPP: ASAPP is an established enterprise AI platform for customer service automation across voice, chat and messaging, with large financial-services deployments. It competes with Gradient Labs in the broader enterprise agent space and is a broader incumbent with a wider industry footprint.

Emerging players

• Hume AI: Hume AI builds emotionally intelligent voice and conversational AI, with overlapping capability in natural, multi-accent voice agents. It is an emerging player whose voice quality and emotion-sensing technology could compete with Gradient Labs' Voice AI Agent over time.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

• Xx.com/GradientLabsAI
• YouTubeyoutube.com/@gradient-labs-ai
• LinkedInlinkedin.com/company/gradientlabs

Compliance10 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles10 records

Subsidiaries1 record

Funding overview

Funding rounds3 records

Investors8 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 Gradient Labs