Frequently Asked Questions (FAQ)

Both technical and non-technical teams who want to move fast when it comes to building AI Agents and Agentic workflows. Sapience helps you fuse together your proprietary data and line-of-business systems, with the latest AI models from OpenAI, Anthropic, Perplexity and Google to create powerful AI systems, purpose-built for business use (unlike other agent platforms that are toys, and built for consumers).

Sapience is designed to help you go from having an idea for an AI system to having a working MVP in hours or days, instead of months.

Then after you've gathered feedback from your early users, to take it to production in days.

The system is designed to rapidly demonstrate value for senior users. If you're a CFO, COO, CEO, CIO, CMO who's "AI curious", then get in touch and we'd love to show you what we can co-create with Sapience.

Nope. That’s deliberate. Sapience is designed for busy (and expensive) executives - and can be a game changer for them in terms of personal, team and org productivity. We use the best and most powerful AI models for the job - whatever that job is - and part of doing that is we don’t need to worry about a population of free users that will never pay anything.

So if you’re using Sapience - someone is paying for it. Probably an organization.

There are free guest user roles, so if you’re already on Sapience, you can easily collaborate with others, without worrying about additional cost.

And having said all of that, its not expensive. A team account starts at $500 a month. This is a trivial expense in context given what Sapience can do for you and your org.

Contact your Account Manager, submit a form here in the documents portal, or email developers@sapiencecloud.ai.

Get in touch. And we actually mean that!

We know that AI systems are best when they're tailored to your organisation, your data, and your systems. We specialize in helping you leverage Sapience to get started fast, and we can help bridge the 'last mile' with our team of AI Engineers.

We’re iterating on the core platform quickly, with major releases now shipping monthly. If you’ve got a feature you think the platform needs, then we’d like to hear from you.

Reach out to your Account Manager. We run regular group webinars on how best to use the system as well as monthly “what’s new?” Webinars. We can organize small-group and 1:1 training for you as well. But, there is a wealth of training material here in the Help portal, including some excellent video tutorials. Start there!

If we list them here, then the list will almost immediately get stale! We use the latest and best models from each of Open AI, Anthropic, Perplexity and Google. This is divided into their “workhorse” models (e.g. “gpt-X” from OpenAI, “sonnet-X” from Anthropic) and their “reasoners” (o-series from OpenAI, Opus etc.).

In addition to this we use task-specific AI models for computer vision, audio transcription, real-time voice support and other edge cases.

Bottom line: if you’re using Sapience, you’ve got access to the latest and greatest models from all 3 of the world’s top AI labs, and you’ll always be on the bleeding edge.

Contact your Org Admin if you want to create a new company-wide Agent. However, there is a good chance that what you want to do can be done within a Project Agent, which all users can create. Projects are Agents! So, if you want an Agent that’s expert on a series of documents (as an example), create a project, describe the knowledge corpus, and fire up the Project Agent Chat. There - you’ve created your first custom RAG agent!

Sapience has its own skills system, and we got there 14 months before Claude Code did! You can import Claude Code skills into Sapience, but the Sapience skills system is similar but more powerful.

There are a series of out-of-the-box Global Skills you can leverage, and you can define your own skills at both the Org and User level.

All the power of a dynamic Agent Skills platform, with flexible scoping. I.e. power + simplicity, which is what we aim for.

Sapience can even port your Claude Code skills.md files INTO a Sapience skill. How’s that for great?

Yes! Sapience has a rich set of automation primitives, including sophisticated ability to connect to any external API.

Even better, we’ve got some dedicated walkthrough’s and how-to guides that explain how to use Sapience with Zapier, Make.com and n8n.

For power users and builders, Sapience can consume any REST API, using GET, POST, PUT, PATCH, and DELETE http verbs.

There are essentially two ways to use Zapience in an automation context:

1) use it as a worker node within one of your automations (this works great) or

2) use Sapience to coordinate your other automations.

We recommend the latter. Wire up your automations (e.g. Enrich-This-Contact-In-Hubspot), and then tell Sapience to do it. Or create a Scheduled Job and have Sapience do that when you want: e.g. “Sapience, I want you to do work every day at 12 pm, get the new form fills from our website using the Get-New-Leads skill, and then enrich them with Clay and store the data in Hubspot”.

If you want to use the first approach, you’ll need access to Sapienza’s API (which is awsome 🙂). Reach out to developers@sapiencecloud.ai

Firstly, “hallucinate” is a friendly word for an extremely dangerous phenomenon with generative AI: it will lie to you.

Sapience won’t. It is designed with proprietary anti-hallucination technology that fights this bug in LLM architecture, meaning that the answers Sapience gives you will always be grounded in facts and sources. Does that mean the answer it gives you will be perfect? No - it might rely on a website that is complete rubbish. But, it will show you exactly where it got its data from. Its also trained with a vast amount of “world knowledge” that goes beyond what LLMs build into their core models, and this helps it find and fight sources that have bad data in them. Read this article to learn more:

Too many to list! Sapience has out of the box integrations with hundreds of different systems ranging from enterprise (Oracle, Salesforce) to consumer-oriented (Notion, Dropbox). Sapience is at its heart a data + AI platform, so integration is one of its strongest points. Whilst the list continues to grow and expand, you can see a list of the big systems and targets that we are regularly asked to integrate with here:

Yes. Every part of the Sapience Platform is exposed as a JSON API using the OpenAPI / Swagger standard.

If you're a Sapience Enterprise customer you get API keys by default.

If you're a developer and you'd like to build against our platform and integrate it with your system, please get in-touch with us at developers@sapiencecloud.ai.

In addition to the API, we use Stainless to create SDKs in popular languages (Typescript, Python today, with C# and Java coming next). Stainless is the company that creates OpenAI's client-side SDKs, so if you've ever worked with OpenAI's client libraries, you'll be right at home using the Sapience SDKs.

Sapience can be deployed in your private cloud, in a dedicated instance managed by us but isolated to your org, or in our managed deployments. We have deployments in the USA, EU and both the UAE and KSA in the Middle East. We have customer deployments on 5 continents.

Sapience follows HIPAA standards for securing data - the same rules that apply to hospitals storing patient data in the USA. It’s the gold standard for data security. Data is encrypted at rest and in transit, with optional private-key encryption available and dedicated deployments. If you need an in-depth view of how Sapience treats data, refer to this article:

Absolutely not. Your data is your data. We don’t share it or train with it or in any other way leak data out to third parties. See article referred to above for an in-depth guide on data management and security/compliance.

Yes! The entire system is built around a series of cloud agnostic deployable containers, that can run anywhere.

We have deployed and hosted on both of AWS and Azure to giver managed customers a choice of cloud environment out-of-the-box.

But if you have custom deployment requirements, or are interested in running this on your own private cloud environment, please get in touch with our Enterprise team and we can talk about it. The system can run on any Kubernetes cluster, although note that it generally has an external dependency to one of the public cloud AI platforms for Generation (e.g. OpenAI, Anthropic, Google, Perplexity).

Data sovereignty and private-cloud support are 1st class citizens in our architecture, and we generally take a “your cloud, our cloud, any cloud” approach.

Absolutely! The Sapience platform enables the rapid creation of custom AI + Data solutions, and is designed from the ground-up to be customized, configured and rebranded.

If you're building apps for your customers or want to create a product offering that combines data and AI to do something new, we'd love to talk to you.

We're product builders at heart and would love to talk about your business and your customers together.