Case study · Legal services
Al Tamimi HR Bot
An NLP assistant that lives inside Microsoft Teams.
- Client
- Al Tamimi & Co.
- Service
- Design and development
- Sector
- Legal services
- Delivered in
- Microsoft Teams
At a glance
What the suite changed
Answers arrive in the tool people already have open, not in one more portal to learn.
Retrieval over the firm's own documents, then generation, so an answer stays tied to source text.
A specialised summariser returns the part of a policy that applies, rather than the document.
The old path was raise a ticket and wait for someone to read the policy for you.
The brief
A law firm with its own knowledge problem
Al Tamimi & Co. is the largest law firm in the Middle East. A firm that reads documents for a living still has the same internal problem everyone else does: corporate policy lives in long documents, and finding the clause that applies to your situation means either reading the whole thing or asking a person and waiting for them to read it for you.
Neither is a good use of anyone's time, and the delay is the actual cost. The question is usually small. The wait is not.
The answer already exists in a document the firm owns. The work is getting it to the person who asked, in the place they asked.
How it works
Ask in Teams, answer from the firm's own documents
Ask
An employee asks in natural language, in the Teams thread they are already in.
Retrieve
A RAG model searches the firm's own policy documents for the passages that actually bear on the question.
Summarise
The summariser condenses those passages to the sections that matter, rather than returning a file to read.
Answer
The response comes back in the same conversation, grounded in the retrieved text rather than invented around it.
The build
Two decisions that carry the product
Retrieval before generation. The suite blends information retrieval with generative text synthesis rather than relying on a model's memory. That is what keeps an answer accurate and tied to the specific context it was asked in, which for a law firm answering its own people about its own policy is the whole requirement.
Teams rather than a portal. An internal tool competes with the habit of asking a colleague. A new web app loses that competition. Delivering the assistant as an interactive Microsoft Teams bot puts it where the question was already being typed, which is the difference between a tool that gets used and one that gets announced.
Wow Labz designed and built the suite: the NLP and retrieval layer, the summarisation of long policy documents, and the Teams bot people actually talk to.
Stack
What it was built with
Questions
The HR Bot, answered
What is the Al Tamimi HR Bot?
A document intelligence suite Wow Labz built for Al Tamimi & Co., the largest law firm in the Middle East. It summarises the sections that matter inside long policy documents and answers employee questions in natural language, delivered as an interactive Microsoft Teams bot.
Why Microsoft Teams rather than a web portal?
Because that is where the question was already being asked. An internal knowledge tool competes with the habit of messaging a colleague, so it has to live in the same place. People get corporate information where they already work instead of raising a ticket and waiting.
How does it avoid inventing an answer?
It retrieves before it generates. A RAG model finds the relevant passages in the firm's own documents first, and the response is synthesised from those. Blending retrieval with generation is what keeps answers accurate and tailored to the specific context rather than plausible-sounding.
What did Wow Labz build?
Design and development of the whole suite: the NLP and retrieval layer, the specialised summariser for critical document sections, and the interactive Microsoft Teams bot that delivers it.
Let's talk
Got a developer product nobody finishes integrating?
Tell us what you are building. We will tell you how we would ship it.
