Read emails and extract information
Identify requests and turn their content into fields such as customer, product, date or amount. Incomplete details are flagged for review.
AI agents that answer questions and carry out tasks
Does your team answer the same questions or copy data from emails? We build agents that consult your knowledge, extract information and act in your systems, with clear rules and review where needed.
Where does support or daily work get stuck?
We choose a specific process and define what the agent handles and when your team steps in.
The same questions keep coming back
Prices, policies and FAQs are spread across people. The assistant searches your sources to prepare an answer.
Useful data arrives inside emails
Orders or requests need to be read and copied. The agent extracts fields, checks what is missing and prepares the record.
Each question opens several tools
An answer requires an order, customer record or ticket. We connect those lookups to the authorised systems.
Exceptions need context
When a request needs human judgement, the handoff includes the conversation and collected details so your team can continue.
Tasks your agent can handle
Look up and record information
Find an order or create a request through an API, the connection between applications, using agreed permissions and checks.
See integrationsAnswer using your business knowledge
Search documents, catalogues or FAQs before answering from that information. This retrieval of sources is known as RAG.
Receive enquiries and prepare opportunities
Capture the need and details a person chooses to share, prepare a record and connect it to the agreed sales follow-up.
Provide real-time voice assistance
Build a spoken conversation connected to business knowledge. For Tremus, we developed this assistance alongside its ecommerce website and app.
Deliver an agent you can evaluate
Tests using cases from your process, error and usage logs, and rules for handing off when information or authorisation is missing.
Tremus: voice assistance with business knowledge
Explore Tremus’s ecommerce website, app and assistant, alongside our internal document retrieval demonstration.
Tremus: ecommerce, an app and voice assistance
An online store, an iPhone app and an assistant that answers in real time by voice using the brand’s knowledge.
View project
AI that connects questions to your business knowledge
An example of source-backed answers, knowledge retrieval and continuity of support.
Explore the workWhat a pilot includes and what it does not
We test a task from your operation, document its behaviour and agree what is needed to put it into production.
What it includes
- Definition of the task, the sources and the permitted actions
- Agent implementation over a bounded process
- Evaluation set with real cases and documented results
- Handoff rules to a person and conversation logging
- A continuation recommendation with estimated running costs
What it excludes
- Model usage, billed by the provider and varying with volume
- Cleaning and rewriting the source knowledge, if needed
- New integrations with systems that do not expose an API yet
- Guaranteed commercial outcomes or error-free answers: quality is evaluated against agreed cases
The pilot is quoted by scope once the task and sources are defined. Build, hosting and model usage are presented as three separate lines so you can compare them.
Recurring third-party costs: Model usage per conversation, hosting and infrastructure, and channel licences (the WhatsApp API, for example).
Can the agent be wrong?
Models can generate incorrect answers. We bound their actions, measure behavior with an evaluation set and hand off to a person when support is insufficient. We agree on indicators and a baseline with you to evaluate quality.
What does it cost to run each month?
It depends on conversation volume and the model chosen. We estimate it before the pilot using your real volume and present it separately from development: build, hosting and usage are three distinct lines.
Do I need my documentation in order?
It helps a great deal, but it need not be perfect. Part of the pilot is finding out what is missing. What is essential is that it exists and that someone owns keeping it current: an agent over stale documentation repeats errors very confidently.
Can it read emails and create records automatically?
Yes, when the scope and access allow it. We define which mailbox it can read, which fields it extracts and where it records them. We validate fields, control duplicates and send exceptions or actions requiring approval for review.
Related services
AI chatbots that put your knowledge to work
Give customers useful answers grounded in your business information. We build assistants with source retrieval, connected tools and a clear handoff to your team.
Connect your systems and stop copying data by hand
An order, enquiry or stock change should not require updating several tools by hand. We connect your applications so data reaches the right place and you can see what happened.
Custom software for the way your business works
If your operation relies on spreadsheets, emails and repeated tasks, we turn that process into a tool of your own. We build the data, screens and connections your team needs to work.
Start with a task that takes up your team’s time
Tell us what comes in, what needs to happen next and which tools you use. We define a first agent, its connections and how to evaluate its work.