AI assistant
Built for lawyers.

AI
for lawyers

We have built a robust AI assistant for South African legal practitioners.
Designed to do all the heavy lifting in your workflow. Your assistant will traverse the breadth of law and pinpoint what matters most.

Technology Stack & Integrations

Free trial!
Full access to Adjunct product suite.

Try For Free

Get full access.

Get Started Now
Free trial
No credit card required
Cancel anytime

See for yourself

Ask us about any matter you're dealing with, and we'll generate and send you a custom report.

Delivered directly to your inbox.

Case Law

Legislation

Ask Adjunct

Your submission has been received and is on its way to your inbox.
Oops! Something went wrong while submitting the form.
the
Company
Amplify
human
Output.
Ambition.
Ingenuity.

Adjunct/ˈadʒʌŋ(k)t/(n.) Something added to another as a supplementary rather than an essential part to that other.  
(adj.) Connected or added to something.  


At Adjunct we believe that technology serves to extend human capability. AI is simply the latest iteration of that. Our mission is to build tools that amplify the abilities of knowledge workers. To start, we have built a sophisticated AI product tailored for South African legal practitioners. This embodies our principles of complementarity, trust and personalisation. Lawyers who use our product will have more capacity for creative problem solving - opening more opportunity to deeply focus on the intricacies of South African law.

meet the
founders

We have embarked on our mission together using our rich blend of skills and experience across AI engineering, business strategy, operations, and product development. This unique fusion of technological acumen and insight into professional services positions us ideally to identify and address specific challenges faced by knowledge workers, starting with legal practitioners in South Africa.

Julita
Ramsunder

COO

Chief Operating Officer

Julita is an economist with a background in strategic and niche economic consultancy. She began her career at a major consultancy building experience in strategic problem solving and intelligent automation then shifted into economic research where she has led and delivered multiple projects in monitoring and evaluation, trade and industry and public economics.  

Alessandro Scarcella

CEO/CTO

Chief Executive Officer

Chief Technology Officer

Alessandro is an experienced AI engineer who studied language models during his postgraduate research. Over the last decade, he has developed AI solutions in a broad range of industries including retail, financial services and online marketplaces. The application of AI to augment human capability and improve society is a source of continuous motivation and purpose.  

Liam
Peterson

CGO

Chief Growth Officer

Liam is a brand strategist and performance marketer, with over 10 years of experience in guiding businesses and global brands to success. His expertise lies at the intersection of marketing and product development, where he seamlessly integrates technology and user experience to drive impactful results.

Our
technology
Tools on the
Frontier.

Adjunct is dedicated to catalysing the potential of AI for the legal profession, ensuring the advantages are harnessed while guarding against common AI pitfalls, such as inaccuracies. In order to achieve this, we make use of a battery of technologies and techniques including vector embeddings, retrieval augmented generation, hybrid search (semantic and keyword), prompt engineering and LLM finetuning.

While every feature has its own unique architecture and considerations, at the core, all our offerings have two ideas that drive our thinking: Verifiability and Complementarity.

1

User
Query

User inputs a free format legal query. Queries can be full questions, bullet points or simply shorthand notes from a client meeting.

2

Query
Refinement

AI assistant will ask follow up questions to refine the query to get more relevant results.

3

Source Document
Search

Taking this refined query, we will search for relevant source documents across public databases, and your private data repository.

4

AI Answer
Engine

Using the retrieved source documents, we will send the refined query with the retrieved source documents to our fine tuned AI models, which will answer the query using the source documents as the basis for every answer.

5

Answer With
Citations

The contextualised reference answer is returned to the user with a set of references so that every aspect can be verified and used for further exploration.

Read Our Blog

Here, we publish in-depth research pieces that dive into the latest AI trends, opinion pieces that critically examine the impact of technology on the legal landscape, and ethics-driven discussions that address the responsibilities and challenges AI brings to industry.

Product
04/09/24
Alessandro Scarcella
The Adjunct Thesis

Adjunct was conceived to amplify human expertise. We set out to build a company that increases the amount of high-quality work in the world. Simultaneously we wanted to make work less mundane & laborious and allow for more creative expression.

Scale your workflow with Adjunct

Join legal practitioners who have streamlined the way they manage their workflow

Frequently asked questions

Have a different question?
Contact us
What can Adjunct help me with?

Adjunct is built to help with three areas of legal work:

  1. Research

  2. Drafting
  3. Compliance
Does the subscription include access to all Adjunct tools and future updates?

Yes, the Adjunct subscription is a single price that is all inclusive. Paid subscribers get access to all current and future products.

Does Adjunct train on user data?

Adjunct does not use any user data to train models or improve the product. The security of sensitive user data is a top priority for us and training models on user data would contradict that.

Will my query archive be available to other Adjunct users?

No, your query archive will only be available & accessible to you.

Is my data secure?

Adjunct uses industry grade encryption (both in transit and rest) built with Amazon Web Serivces's trusted platform.

How do I know I can trust Adjunct's outputs?

While all generative AI outputs should be double checked, Adjuncts outputs are:

  1. Rooted in primary legal source documents

  2. Accompanied by references back to the source documents
What data sources does Adjunct have access to?

While all generative AI outputs should be double checked, Adjuncts outputs are:

  • National Legislation

  • Case Law

Coming Soon

  • Provincial Legislation

  • Regulation

  • Municipal Bylaws