Built to complete the job. Not just the conversation.
Answers about Stubber's fully autonomous agents, enterprise deployment, integrations, security and transparent utilisation-based pricing.
About Stubber & Why Choose Us
What does Stubber do?
Stubber builds fully autonomous agents for enterprise businesses, with a particular focus on customer sales and support. These agents do more than answer questions: they engage customers, use business systems, take actions and work to complete an entire job from start to finish. Human escalation is available when necessary, just as an employee can ask a colleague for assistance, but it is not the intended outcome of every interaction. We build agents that complete jobs, not just conversations.
What makes Stubber different?
Stubber combines a purpose-built platform with experienced solution architects who design, deploy and optimise fully autonomous agents every day. The platform has evolved through years of live operational use, with lessons from more than 9 million jobs processed translated into capabilities, integrations and safeguards. Customers benefit from that foundation rather than having to discover and solve each challenge independently.
Why choose Stubber instead of experimenting with autonomous agents ourselves?
Autonomy comes with known challenges and unknown unknowns—issues that often only appear when agents operate at volume, across real business systems and with real customers. A limited prototype can be relatively quick to build; an agent that works reliably in production is much harder. Stubber brings years of hands-on experience and a platform shaped by real deployments, helping businesses avoid unnecessary trial and error and reduce the risk of a disappointing first rollout.
Is Stubber proven in real-world environments?
Yes. Stubber has agents operating in live business environments, executing customer-facing processes rather than simply demonstrating what might be possible. The platform has processed more than 9 million jobs. That figure refers to jobs processed—not individual chat messages or a claim that every job resulted in a successful outcome. Our emphasis is on operational performance, continuous improvement and measurable business value.
Why are dedicated solution architects important?
Fully autonomous agents are not best approached as an occasional side project. They require detailed process understanding, integration expertise, engineering judgement and regular optimisation as models and technologies evolve. Stubber's solution architects specialise in this work every day and bring experience from different business environments. That dedicated focus helps identify problems earlier and accelerate the path to dependable production use.
Capabilities & Use Cases
What is a Stub?
A Stub is a job, not a single message. It represents a business task or process undertaken by an autonomous agent, potentially involving multiple messages, decisions, integrations and actions as the agent works toward an outcome. For example, a vehicle-sales enquiry might involve discovering the customer's needs, identifying suitable vehicles, collecting information and progressing a finance application—all within one job. This is why Stubber measures jobs processed, rather than simply counting conversations or messages.
Can Stubber build an agent for any business process?
Stubber can support a wide range of processes, but our core specialisation is fully autonomous customer sales and support, particularly via WhatsApp and web chat. These are areas where we have deep experience and mature integrations. If your business has other high-volume, repeatable processes, our team can assess their technical feasibility, integration needs and commercial value. Not every process that can be automated should be automated.
Is Stubber a plug-and-play solution?
No. Stubber provides a proven underlying platform, but the agent itself is tailored to your business. Our solution architects work with your processes, systems, rules and customer expectations to build an agent that reflects how your organisation operates and what makes it distinctive. You gain reusable platform capabilities without being forced into a generic, one-size-fits-all workflow.
Can Stubber work with systems that do not have APIs?
Often, yes. APIs are preferred, but they are not the only integration route. Stubber has existing integrations with a range of business systems and databases, and our built-in Virtual Worker can interact with applications through their user interfaces where appropriate. This may provide an interim or longer-term route to automation. Our architects assess reliability, access permissions and security before recommending an approach; not every system will be suitable for the same method.
What is the Stubber Concierge Agent?
The Concierge Agent is designed to provide continuity and personalisation across sales and support interactions. Where authorised and configured, it can use relevant customer history, preferences and prior interactions so people do not have to start from scratch each time. It combines that familiarity with the ability to take action through connected business systems and complete jobs. Data access and memory remain subject to the customer's agreed security and privacy controls.
Pricing & Commercial Model
How much does Stubber cost?
There are two main cost components. Platform utilisation is measured through Stubbucks, our prepaid usage credits, consumed as agents perform jobs. Solution architecture and engineering are generally charged for the work of discovering processes, designing, integrating, building and optimising agents. This approach lets businesses scope an initial use case, understand the engineering investment and estimate ongoing utilisation before scaling. Commercial packages can be discussed with our team.
How do Stubbucks work?
Stubbucks are prepaid credits loaded into your organisation's Stubber account. One Stubbuck costs US$0.10. Each job consumes credits based on its configured components and actual utilisation, which may reflect process complexity, communication channels, integrations and other platform capabilities. Once a use case has been built and tested, we can estimate the cost per job and forecast spend against expected volumes. For example, 5 Stubbucks per job = US$0.50 per job; at 10,000 jobs, that would be US$5,000 in illustrative platform utilisation. Actual costs vary by configuration.
Is there a platform access fee?
No. Stubber does not charge a separate fixed fee simply to access the platform. Our platform revenue is incorporated into usage pricing for the components that deliver your agents' jobs. Solution architecture, engineering and any separately agreed services or operational packages are additional where applicable. This is different from saying that every engagement has no upfront costs or no minimum commitments.
How does Stubber help manage LLM and agent costs?
Our platform has built-in guardrails and optimisation mechanisms designed to make language-model usage more efficient and costs more predictable. Capabilities such as the Knowledge Library help agents retrieve relevant business information efficiently. Alongside solution-architect optimisation and visibility into expected cost per job, this helps businesses plan and refine utilisation. Exact consumption can vary by job; no platform can guarantee identical costs for every scenario.
Do I need separate contracts with LLM providers or Meta?
Generally, no for the underlying services included in your agreed Stubber job configuration. Stubber manages its relevant provider relationships and includes applicable underlying technology costs in its quoted utilisation pricing. That gives you one consolidated cost for the included platform components. Certain external systems, licences or third-party services outside the agreed configuration may carry their own fees; these should be identified during scoping.
Am I tied into a long-term contract?
Our philosophy is to earn continued usage by delivering value, rather than create unnecessary lock-in. The utilisation model makes it possible to test, measure and scale gradually. However, specific projects, prepaid packages, operational commitments or Results-as-a-Service arrangements may have agreed contractual terms. Your proposal will make those commitments clear before you proceed.
Would Stubber consider Results-as-a-Service?
Yes, for suitable use cases. Alongside our standard usage and engineering model, we may consider a contracted Results-as-a-Service arrangement linked to clearly measurable business outcomes, such as successful sales or completed transactions. Depending on the agreement, Stubber may fund platform utilisation and engineering in return for performance-based fees or commissions. Such engagements require mutually agreed scope, measurement, economics, responsibilities and contractual safeguards, and are assessed case by case.
Getting Started & Deployment
How do we get started with Stubber?
Start by contacting our sales team for an online or face-to-face demonstration. If you see a promising use case, we'll arrange a workshop to unpack your business process, technical needs, system integrations and desired outcomes. From there, we can scope the engineering effort and estimate the build and utilisation costs. We recommend an R&D mindset: start with a manageable use case, test it in a controlled real-world setting, and invest further when the results justify it.
How should autonomous agents be deployed?
Treat deployment much like onboarding a new employee. Begin with limited responsibility—for example, after-hours enquiries, a defined customer segment or a narrow set of tasks. Measure outcomes, identify edge cases and refine the agent. As it demonstrates competence and reliability, increase its scope and volume. Stubber supports this iterative journey: deploy, measure, refine and scale, with human escalation available where required.
Can businesses of any size use Stubber?
Yes, although the platform is designed primarily for enterprise-grade requirements, including high volumes, complex integrations, security and compliance needs. Smaller businesses can benefit too, especially when a process occurs often enough for the operational value to justify the engineering investment. What matters most is the nature, volume and value of the work—not only the size of the company.
Should we build an autonomous agent in-house first?
You are welcome to build your own agents, but we recommend starting the R&D journey with Stubber rather than spending months rediscovering production challenges. An underperforming first attempt can reduce confidence among management, employees and customers. Starting with our platform and solution architects gives you an opportunity to establish a reliable benchmark, learn what full autonomy can achieve and then decide whether to build additional capabilities internally. We are comfortable being compared on real-world performance.
Can our team build agents directly on the Stubber platform?
Yes. You can build on the Stubber platform yourself. Our recommendation is to begin with Stubber's experienced solution architects or qualified implementation partners, so your team can learn from a successful deployment while avoiding common pitfalls. As your organisation develops dedicated expertise, you can take on more of the design, build and optimisation work in-house. Autonomous engineering deserves focused ownership—not a spare-time assignment for an already busy team.
Enterprise Security & Data
What security and compliance measures does Stubber have?
Stubber is ISO 27001 certified, has completed penetration testing and maintains insurance arrangements relevant to its services, as represented by our team. Security considerations form part of agent design and enterprise deployment, particularly where agents interact with customer information and operational systems. Our team can provide available supporting information and documentation during procurement and security reviews. Specific certification scope, test dates and insurance details should be confirmed through the current documentation.
Will the underlying LLMs train on our private enterprise data?
No. Stubber uses enterprise arrangements with its LLM providers under which customer conversations and private business data are not used to train the underlying models. Agents still process relevant information to perform authorised tasks; processing, retention, privacy and model training are separate considerations. Our team can explain the applicable data-handling arrangements and controls for your deployment.
Ready to explore your first use case?
Speak to the Stubber team to arrange a demonstration and discuss a practical starting point for your business.