Both Gradient Labs and Lorikeet were built for regulated customer support, not generic deflection. The right choice depends on which regulated obligations you carry, which channels your customers actually use, and how you want to pay.
Gradient Labs and Lorikeet are two AI customer support platforms purpose-built for regulated financial services. Both resolve complex tickets end-to-end, run compliance guardrails on every turn, and price on outcomes rather than seats. The difference is in focus: Gradient Labs leads with specialist agents tuned to UK and EU financial regulation, while Lorikeet leads with omnichannel resolution (including sub-1-second voice), deterministic plus natural-language workflows, and defence-in-depth validation across regulated industries beyond banking.
Gradient Labs was founded in 2023 by former Monzo AI leaders and is deployed at banks and fintechs including Wise, Monzo, and Stash, supporting 32M+ end users per the company.
Lorikeet is built for complex and regulated companies (fintech, financial services, healthtech, insurance, gaming), with about 80% of customers being US financial institutions and fintechs.
Both run continuous guardrails: Gradient Labs cites 20+ guardrails on every turn; Lorikeet runs adversarial simulations, inbound message checks, outbound guardrails, and 100% post-facto QA.
Pricing diverges on transparency. Gradient Labs publishes an outcomes-based model with no platform fee but no public rate; Lorikeet publishes per-resolution pricing (about $0.80–$0.95 per chat, email, or SMS; about $1.20–$1.50 per voice) and does not charge for escalations.
Channel coverage is the clearest split: both do email, text, and voice, but Lorikeet also runs WhatsApp and outbound re-engagement on one engine, with sub-1-second voice latency.
Last updated: June 2026
Most AI support vendors sell a deflection rate. Regulated buyers need something else: an agent that can take the correct action on a KYC unlock, a card dispute, or a missed payment, and produce a record their compliance team can sign off on before launch. Gradient Labs and Lorikeet both clear that bar, which is rare. This comparison is written to help a regulated CX or compliance leader decide between two credible options, not to declare a single winner. We are Lorikeet, so we are not neutral, but we have kept Gradient Labs' real strengths on the page because pretending a strong competitor is weak helps no one making a six-figure decision.
Gradient Labs vs Lorikeet at a Glance
Founded and focus. Gradient Labs (2023, London, ex-Monzo founding team) focuses on specialist AI agents for regulated financial services, with a strong UK and EU regulatory lens. Lorikeet (founded 2023) builds AI concierges that resolve issues end-to-end for complex and regulated industries spanning fintech, financial services, healthtech, insurance, and gaming, weighted toward US financial institutions.
Channels. Gradient Labs supports email, text, and voice. Lorikeet supports chat, email, voice (sub-1-second latency, multilingual with automatic language switching), SMS, and WhatsApp, plus outbound voice, SMS, and email re-engagement with compliance controls (DNC, call-hour rules, consent).
Workflows. Gradient Labs uses specialist agents trained on your SOPs and natural-language procedures. Lorikeet combines natural-language workflows and deterministic structured workflows in a single interaction, all configurable in plain English.
Guardrails and validation. Gradient Labs runs 20+ guardrails on every turn with domain-specific compliance coverage. Lorikeet runs defence in depth: pre-launch adversarial simulations and red-teaming, inbound message checks, outbound guardrails, and 100% post-facto QA through its Coach agent.
Compliance posture. Both hold SOC 2 Type 2 and are GDPR-aligned. Gradient Labs emphasizes UK and EU rule coverage (FCA Consumer Duty, CONC, Reg E and Reg Z, PSD2, EU AI Act). Lorikeet is SOC 2, BAA-ready for HIPAA, GDPR-aligned, with PII redaction, RBAC, and data residency in the US, AU, and UK, and has passed security reviews at major US banks. Both support obligations rather than guaranteeing regulatory outcomes.
Pricing. Gradient Labs: outcomes-based, no platform fee, rate not published. Lorikeet: about $0.80–$0.95 per chat, email, or SMS resolution, about $1.20–$1.50 per voice, Coach about $0.25–$0.30 per ticket, escalations not charged, customer defines what counts as a resolution.
What Both Platforms Get Right for Regulated Support
It is worth being clear about what these two have in common, because it is most of the hard part. Both are agentic, not retrieval-and-reply chatbots. Both execute multi-step actions inside your systems rather than handing the customer a help-center link. Both run compliance guardrails continuously instead of treating safety as a post-hoc filter. Both price on outcomes, which aligns the vendor's incentive with resolved tickets instead of seats sold. And both have real regulated customers in production rather than pilots, which is the single best signal that the architecture survives contact with a compliance team.
If you are choosing between Gradient Labs and Lorikeet, you have already filtered out the chatbot vendors wearing an agent t-shirt. The remaining decision is about fit, not category.
Where Gradient Labs Is Strong
Gradient Labs has genuine strengths that matter for the right buyer, and a fair comparison has to name them.
Deep UK and EU regulatory specialization
Gradient Labs was built by people who ran AI at Monzo, and that shows in the regulatory depth. The platform's guardrails are tuned to UK and EU frameworks (FCA Consumer Duty, CONC, Reg E and Reg Z, PSD2, and the EU AI Act) at a level of specificity that a US-first vendor may not match out of the box. For a UK or EU bank where Consumer Duty evidence is a board-level concern, that domain tuning is a real advantage.
Specialist agents for distinct financial workflows
Rather than one general agent, Gradient Labs ships purpose-built agents for lending (missed payments through collections and repayment plans), disputes (intake through chargeback), and KYB identity and document verification. If your support load is concentrated in those exact workflows, starting from a specialist agent can shorten time to value versus building from a blank workflow canvas.
Proven scale at established banks
Deployment across institutions including Wise, Monzo, Stash, Current, and Rho, supporting 32M+ end users and reporting resolution rates of 80-90% with CSAT scores that the company says exceed human teams, is a strong production track record. For a buyer who weights logos and scale heavily, that portfolio carries weight.
Simple, transparent pricing philosophy
No platform fee and pay-only-for-successful-resolutions is a clean model that is easy to explain to a CFO. The trade-off is that the per-resolution rate is not published, so total cost still requires a sales conversation.
Where Lorikeet Differentiates
Lorikeet and Gradient Labs overlap on the regulated-agent thesis, but they diverge on channel breadth, workflow model, validation depth, and pricing transparency.
Omnichannel on one engine, including sub-1-second voice
Lorikeet runs chat, email, voice, SMS, and WhatsApp as the same agent on one workflow engine, with shared context across channels, plus outbound re-engagement. Voice runs at sub-1-second latency with multilingual support and automatic language switching. Gradient Labs covers email, text, and voice; Lorikeet adds WhatsApp and outbound, which matters if your customers reach you across more than three channels or you run proactive collections and abandonment flows. When a customer starts in chat and calls in, the same agent picks up rather than starting over.
Deterministic and natural-language workflows in one interaction
Some regulated steps must happen in an exact order every time (disclosures, identity checks, dollar-threshold approvals), while others benefit from open-ended reasoning. Lorikeet lets you combine deterministic structured workflows with natural-language workflows inside a single interaction, so you get scripted certainty where the regulator demands it and flexibility everywhere else. This is useful when you want provable, repeatable behavior on the regulated path without losing conversational quality on the rest.
Defence in depth with 100% automated QA
Both platforms run runtime guardrails. Lorikeet adds two bookends most vendors skip. Before launch, it runs adversarial simulations and red-teaming so your compliance team can read pass and fail results before go-live rather than approving on faith. After every interaction, the Coach agent runs 100% post-facto QA (root-cause analysis, ticket quality score, resolution verification), which is the AI evaluating the AI on every ticket rather than a sampled audit. Coach is also deployable standalone at about $0.25–$0.30 per ticket if you want QA on your existing human or AI volume first.
Published per-resolution pricing and a customer veto on resolutions
Lorikeet publishes its rates: about $0.80–$0.95 per chat, email, or SMS resolution, about $1.20–$1.50 per voice, escalations not charged, and the customer holds the veto on what counts as a resolution. That last point matters for an outcomes model: if the vendor defines resolution, the incentive can drift toward counting marginal outcomes. Letting the customer define it removes that ambiguity. Gradient Labs uses outcomes-based pricing with no platform fee, which is also customer-friendly, but the rate is set in a sales conversation rather than published.
Breadth beyond banking
Lorikeet serves healthtech (BAA-ready for HIPAA), insurance, and gaming alongside fintech. If your regulated footprint spans more than UK and EU banking, that breadth and the US-bank security pedigree may fit better. Gradient Labs' depth in UK and EU financial regulation is the flip side of that trade-off.
Audit Trails and Regulator Examinations
For a regulated business, the audit trail is the deliverable that matters most. When a KYC unlock or a dispute decision is questioned months later, you need to point at the exact reasoning step and tool call where the agent acted, not hand over a chat transcript and hope it is enough. Both Gradient Labs and Lorikeet are built with this in mind. Gradient Labs cites comprehensive audit logs, SSO, and role-based permissions as part of its banking-grade data handling. Lorikeet's record is designed for compliance approval before go-live and for regulator examinations after, with the Coach agent verifying resolution and assigning a ticket quality score on every interaction rather than on a sample. The practical question to put to either vendor is the same: can you replay the full reasoning chain plus tool calls for any ticket from 90 days ago, in order, with timestamps. A sampled log is not the same artifact a regulator wants, and the distinction is easy to miss in a polished demo.
Integrations and Implementation
Both platforms expect to plug into the systems you already run. Gradient Labs connects via API, CSV upload, or inside your existing tools, with specialist agents trained on your SOPs and your best human agents' procedures. Lorikeet integrates across ticketing (Zendesk, Intercom, Front, Kustomer), CRM and telephony (Salesforce, including coexistence with Agentforce, Talkdesk, Twilio, Amazon Connect, Aircall), and knowledge sources (Notion, Confluence, Google Drive, Guru), with least-privilege scoped tools and webhooks, plus a Lori MCP server for Claude Code and ChatGPT. On implementation, Lorikeet pairs you with a forward-deployed PM and engineer, stands up a sandbox in 20 to 30 minutes, and typically reaches production in about a month. Both vendors are agentic platforms that require configuration and a compliance review, so neither is a self-serve, two-week chatbot. The question to ask is how much of the workflow your own team can own after launch versus how much depends on the vendor's embedded staff, because that determines your cost and agility a year in.
LLM Architecture and Reliability
Reliability in a regulated setting depends partly on not being a single point of failure. Gradient Labs runs multi-model infrastructure across OpenAI, Anthropic, and Google with failover across cloud providers. Lorikeet similarly switches dynamically between Anthropic, OpenAI, and Gemini by task, and holds contractual no-train agreements with those providers so your customer data is not used to train third-party models. For a compliance team, the no-train commitments and the multi-provider failover are both worth confirming in writing, because they speak to data governance and uptime respectively, and the answers should be specific rather than reassuring.
An Honest Look at Lorikeet's Limitations
No platform is the right answer for every buyer, and Lorikeet is no exception. If your support volume is concentrated entirely in UK or EU lending, disputes, and KYB, and you want to start from an agent pre-tuned to those exact workflows and to FCA and EU AI Act specifics, Gradient Labs' specialist-agent approach may get you to first value faster than configuring workflows from scratch. Lorikeet's center of gravity is US financial services, so a UK or EU-only bank should pressure-test the regulatory tuning against its specific obligations during the sandbox. And both vendors require a real implementation: Lorikeet uses a forward-deployed PM and engineer, with a sandbox in 20-30 minutes and typical production in about a month, which is fast for the category but is not a self-serve drop-in. If you want a two-week chatbot, neither of these platforms is what you are shopping for.
How to Choose Between Gradient Labs and Lorikeet
The decision usually comes down to four questions.
Which regulatory regime dominates your obligations?
If you are a UK or EU bank where FCA Consumer Duty and EU AI Act evidence is the central concern, Gradient Labs' specialization is a strong fit. If you are a US fintech or financial institution, or you span healthtech, insurance, or gaming, Lorikeet's US-bank pedigree, HIPAA readiness, and multi-industry breadth align more directly.
How many channels do your customers actually use?
If email, text, and voice cover your volume, both platforms fit. If you need WhatsApp, or you run outbound collections and abandonment re-engagement, Lorikeet's omnichannel-on-one-engine model with outbound is the broader fit.
Do you want scripted certainty on the regulated path?
If your compliance team needs specific steps to happen in an exact order every time, ask both vendors how they enforce that. Lorikeet's deterministic structured workflows are designed for exactly this, combined with natural-language reasoning elsewhere.
How much pricing transparency do you need up front?
If you want a published per-resolution number you can model before a sales call, Lorikeet lists its rates. If you are comfortable negotiating an outcomes rate and value no platform fee, Gradient Labs' model is clean and competitive.
Questions to ask both vendors
Show me an end-to-end record of a regulated decision your agent made last week, with every tool call and the reasoning between them.
Can my compliance team run your guardrail and simulation suite before go-live and read the pass and fail report?
What happens when a downstream system returns a 5xx mid-workflow: retry, escalate, or roll back?
Who defines what counts as a resolution for billing, you or me?
Is voice on the same engine as chat and email, and can the agent take regulated actions on a call?
Which of my exact regulatory frameworks are your guardrails tuned to today, and which would need configuration?
Lorikeet's Take
Gradient Labs is a serious platform, and a UK or EU bank concentrated in lending, disputes, and KYB should shortlist it. The reason to choose Lorikeet is not that the competitor is weak. It is fit: if your regulated footprint is broad (US financial services, healthtech, insurance, gaming), if your customers reach you across chat, email, voice, SMS, and WhatsApp plus outbound, if you need deterministic certainty on the regulated path combined with natural-language flexibility, and if you want published pricing with a customer-held veto on what counts as a resolution, Lorikeet is built for that shape of problem. The honest test for either vendor is the same: can your compliance team sign off on the behavior before launch, and is the agent correct on the hard tickets rather than only the easy ones.
Key Takeaways
Gradient Labs and Lorikeet are both genuine regulated-support platforms; the choice is about fit, not category.
Gradient Labs is strongest for UK and EU banks wanting specialist agents pre-tuned to FCA Consumer Duty, CONC, PSD2, and the EU AI Act, with a proven portfolio at Wise, Monzo, and Stash.
Lorikeet differentiates on omnichannel (chat, email, sub-1-second voice, SMS, WhatsApp, plus outbound), combined deterministic and natural-language workflows, defence-in-depth validation with 100% QA, and published per-resolution pricing.
Both hold SOC 2 Type 2 and are GDPR-aligned; both support compliance obligations rather than guaranteeing regulatory outcomes.
Pressure-test either platform with your hardest regulated tickets and a pre-go-live compliance review before signing.
If you are weighing Gradient Labs against Lorikeet, book a Lorikeet demo and bring your hardest regulated tickets; we will run them against your guardrails in a sandbox before you commit.
