An AI knowledge base, powered by SmartRetriever
, our own product. It is vector and graph RAG, with hybrid retrieval. Answers come from your own documents and systems, cited and permission aware, with the source shown.
At least 3x improvement on the process we take on, with our fee tied to it.
SmartRetriever, our own product, answering from your own documents and systems, with the source shown for every answer and only what each person may see.
In your own Azure, with models accessed through Microsoft Foundry. Your data is never used to train models.
On how long it takes your people to find a trusted answer, or the questions a team resolves per person, against a baseline agreed first. Our fee is tied to at least 3x.
SmartRetriever is our own product. We connect your first sources, set it up for your people and measure how much faster they find a trusted answer.
Staff ask the same questions again and again, and the answers live in the heads of a few people. When one of them leaves, the answer goes too.
A public chat tool does not know your documents, and pasting them into it raises its own questions about where they go.
An AI knowledge base powered by SmartRetriever, our own product, with vector and graph RAG, cited answers and permission-aware access.
SmartRetriever is the engine behind this service. It is our own product, built by Sendient.
The SmartRetriever chat is on this page. Use the chat button at the bottom right and ask it about Sendient.
See SmartRetriever working →SmartRetriever is our ingestion and retrieval engine as a product. It combines vector, keyword and graph retrieval so each answer draws on the right context for the person asking. It starts as a website assistant and grows into the core of how your people find things, with the same permissions and audit trail at every step.
Hybrid retrieval by meaning, by exact terms and by the connections between documents.
People only get answers from documents they are already allowed to open. Permissions come from your existing systems.
The right context for each person, department, location and country, every time, with security and privacy built in.
SharePoint, OneDrive and Google Docs, and the systems behind your processes, kept in step so answers stay current.
A WordPress plugin or web bot puts it on your website, and it can run inside your own processes.
Multi-step answers that follow cross-references across documents.
Every answer points to the page, the region on the page or the moment in a recording, with a note when sources disagree.
PDFs, scans, Office documents, images, audio and video, with timestamped transcripts.
Checks across documents that find where they contradict each other or disagree.
A remote MCP server, a REST API and Python and Node SDKs, so your AI tools and apps can query it.
A per-project mode that keeps processing in Azure only, for sensitive content.
Your documents, questions and answers are never used to train any model, ours or anyone else’s. Data is encrypted in transit and at rest, you choose where it is processed, and every query and answer is logged.
RAG means retrieval augmented generation. Instead of asking a model to remember, we find the right passages in your own documents first, then ask the model to answer from them.
Finds passages by meaning, so a question about leave can find a policy about holiday.
Finds exact terms, names and codes that meaning alone can miss.
Follows the connections between documents, people and topics to find the right context.
Before anything is shown, the answer is limited to what that person is already allowed to open.
A private LLM means the model works inside your own cloud, under your controls, with your data staying there. We route each task to the best model for the requirement, from OpenAI, Anthropic and Google and open-weight models, accessed securely through Microsoft Foundry on Azure.
We never train models on your data. What you share stays yours.
The links between documents, teams and topics are kept, so related material turns up together.
One place to ask across SharePoint, OneDrive, Google Docs, a website plugin and the systems behind your processes.
The index is kept in step with your sources, so answers use the current version.
The documents and systems that matter first, brought into one engine.
Each answer shows the passages it used, so people can check it.
People only get answers from what they are already allowed to open.
A record of who asked what, and what was searched.
We connect the first sources and map who can see what.
Documents are split into passages and indexed by meaning, by exact terms and by connections.
We check the answers against questions your people really ask.
We add people, sources and channels on the same engine.
Run and improve. After go live we keep the work running and keep improving it with you.
For a knowledge base, 3x is measured on how long it takes a person to find a trusted answer, or how many questions a team resolves per person, against the baseline we agree first.
At least 3x improvement on the process we take on, with our fee tied to it.
How we charge →These are Sendient’s own products, and the retrieval behind them is the retrieval we build for you.
Gives answers from your documents and systems, current, cited and permission aware, with every source shown and every query logged.
A website plugin or web bot that answers visitors from your own content and passes warm leads to your team by email.
Shows the exact evidence behind every finding, highlighted on the original, so every outcome can be checked.
Every system we build keeps a person in charge of the outcome, and every action is logged.
Data is classified and protected before any model sees it, and access follows each user.
Built on Microsoft Azure, in Sendient’s tenant or your own, with data kept in-country.
We never train models on your data. What you share stays yours.
Sendient holds Cyber Essentials Plus, ISO 27001, ISO 9001 and Investors in People. See our certifications
SmartRetriever is the engine behind this service.
See SmartRetriever →
Testing, inspection and compliance. Any standard, more than 100 integrations and a mobile app for inspections on site.
See SmartAssessor →
For schools, universities and professional bodies. Marks written and typed work, detects AI, tailors learning and runs oral assessment in Teams.
See SmartEducator →
Agent-led interviews in Teams, CV and job matching on meaning, screening and shortlisting, working around the clock.
See SmartRecruits →Something else on your mind? Ask us directly and a person will reply.
Ask a questionRetrieval augmented generation. The system finds the right passages in your documents first, then the model answers from them and shows its sources.
Models are accessed through Microsoft Foundry in your own Azure, and we never train models on your data, so your documents do not improve anyone else’s model.
No. Access follows the user, with permissions taken from your existing systems.
We agree a measure, such as the time to a trusted answer, and a baseline before we start. The fee is tied to at least 3x on it.
Tell us where your knowledge lives today. A person at Sendient will reply with how a knowledge base would work for you.
At least 3x improvement on the process we take on, with our fee tied to it.
Thank you, . A person at Sendient will reply to your work email.
Your email, calendar, Teams, documents and business systems connected to AI through MCP and APIs, safely.
Read more →
AI built on Microsoft Azure in your own tenancy, hosted in the UK, with models, retrieval and agents measured on one process.
Read more →
A digital colleague on your website that answers customers in their own language, day and night, and hands warm leads to your team.
Read more →At least 3x improvement on the process we take on, with our fee tied to it.
