Welcome to Kallam
This guide walks your team through every feature of the Kallam platform. Each section explains what the feature does, how to use it step by step, and what to keep in mind when working with it. Whether you are an experienced disputes practitioner or new to AI assisted document review, this documentation is designed to give you complete confidence in using Kallam.
Document intelligence, purpose built for disputes.
What Kallam Does
Kallam is an intelligent workspace purpose built for document heavy legal disputes. It transforms raw case files (PDFs, scanned documents, emails, contracts, and more) into structured, searchable, and citeable case intelligence. Instead of spending weeks manually reading, renaming, dating, and organising thousands of documents, your team uploads them to Kallam and the platform handles the heavy lifting automatically.
Kallam was designed specifically for the workflows of international arbitration and commercial litigation. Every feature maps to a real task that disputes lawyers perform daily: building chronologies, searching across evidence, extracting structured data from documents, and managing exhibit citations in briefs.
The End to End Workflow
Working with Kallam follows four stages. Understanding this workflow will help you see how each feature fits together.
- Upload. Drag and drop your case files into a Kallam case. You can upload PDFs, scanned documents, Word files, images, and even ZIP archives containing multiple files. There is no practical limit on the number of files you can upload at once.
- Process. Kallam reads every document automatically. It extracts dates, generates descriptive titles, classifies each document into a category (Contract, Correspondence, Report, etc.), and creates a summary. If your documents are scanned, Kallam applies OCR (optical character recognition) to make them text searchable. Bundled PDFs, where multiple documents have been scanned into a single file, can be split into individual documents before processing.
- Analyse. Once your documents are processed, Kallam's analysis tools become available. You can search across the entire corpus using Finder, ask questions and get cited answers using the Agent, extract structured data using the Tabular Extractor, collect generated work products on the Artifacts page, and build a timeline that links every event back to its source document.
- Deliver. Export your timelines, exhibit tables, and summaries for use in submissions. Use the Word Add-In to manage exhibit numbering and insert citations directly into your briefs without leaving Microsoft Word.
Organization Dashboard
The Organization Dashboard is the workspace where the organisation administrator controls the firm's Kallam subscription. From this dashboard, the admin manages the subscription plan, purchases additional processing capacity, monitors usage across the team, and invites or removes team members.
Subscription Management
The organisation admin selects the subscription tier for the firm. Kallam offers three tiers: Starter, Pro, and Enterprise. The admin can switch between Starter and Pro or upgrade to yearly billing directly from the dashboard. Yearly billing offers a discount compared to monthly payments. The subscription tier determines which features are available to all users in the organisation. For example, the Word Add-In and exhibit numbering are only available on the Pro tier. Each tier also comes with a monthly page allocation per seat (for example, Starter includes 1,000 pages per seat per month and Pro includes 5,000). Enterprise is sold with dedicated customer success and custom pricing. Contact info@kallam.ai for a quote.
Page Credit Add-Ons
Every Kallam subscription includes a monthly page allocation. When your team needs additional processing capacity, the admin can purchase add-on page credit packs directly from the dashboard. These are one time purchases that do not affect the recurring subscription. Three pack sizes are available (small, medium, and large); larger packs carry a lower per page cost. The current pack sizes and prices are shown in the Organization Dashboard, so the values you see there are always live. Credits from add-on packs never expire and are applied after the monthly allowance is consumed.
Analytics
The dashboard provides a clear overview of how Kallam is being used across your organisation. The analytics panel shows:
- Number of Cases created. How many cases currently exist in the organisation.
- Documents uploaded. The total count of documents that have been uploaded across all cases.
- Pages processed. The total number of pages that Kallam has processed, which directly corresponds to page credit consumption.
- Tokens consumed. A measure of how much the Agent and Tabular Extractor features have been used. This helps the admin understand the team's usage of AI powered features.
- Usage per user. A breakdown showing how much each team member has contributed to overall usage, helping the admin identify who is actively using the platform.
Team Management
The organisation admin can add or remove team members at any time. There are two ways a new user can join your organisation:
Option A: Admin Sends an Invitation
The admin clicks "Invite Member" and enters the new team member's email address. Kallam sends an invitation email to that person. The recipient clicks the link in the email and is guided through account setup. Once their account is created, they are automatically added to the organisation.
Option B: User Signs Up and Joins with OTP
Alternatively, the new user can sign up for Kallam through the normal sign up flow on the Kallam website. After creating their account, they click "Join Organization" within their settings. The admin provides them with the one time password (OTP) that was generated when the invitation was created. The user enters this OTP to join the organisation.
If a team member has not received the invitation email, ask them to check their spam folder. Alternatively, use Option B: have them sign up directly and join using the OTP code. Both methods produce the same result. The user becomes a member of your organisation.
Subscription Administration
For non enterprise subscriptions, the organisation admin has full control over the subscription. The admin can add or remove user seats (each seat corresponds to one team member), manage billing information, and view invoices, all without needing to contact Kallam support. Enterprise customers have a dedicated customer success manager for subscription changes.
Case Management
Every dispute in Kallam lives inside a Case. A case is an isolated dataroom, a self contained workspace where your documents, timeline, searches, and AI interactions are kept entirely separate from other cases. This isolation ensures that data from one case never leaks into another, which is critical when handling confidential legal disputes.
Creating a Case
To start working on a new dispute, create a new case from the Home page. Each case acts as an independent dataroom where you will upload all documents related to that specific case. Give your case a clear, descriptive name so your team can easily identify it. For example, "Acme Corp v. Northwind Industries" or "Meridian Holdings, Commercial Arbitration." Once created, the case is ready to receive documents.
Navigating Between Cases
If you are working on multiple cases simultaneously, you can switch between cases at any time. The Home page shows all cases you have access to, with tabs to filter between cases you own and cases shared with you, a search bar, and filters for case type. Clicking on any case opens its dedicated workspace. Your work in one case does not affect any other case. Each is completely independent.
Case Admin and Permissions
The user who creates a case automatically becomes the case admin for that case. The case admin has full control over the case, including the ability to:
- Add team members. Invite colleagues to join the case so they can view documents and use all of Kallam's features within that case. Only organisation members can be added.
- Manage access. Team members added to a case can view all documents, use the Agent, Finder, Timeline, and Tabular Extractor within that case. They cannot see other cases unless they have been explicitly added to those as well.
Case Settings
The case admin can access Case Settings to manage the case configuration. From here, the admin can:
- Rename the case and edit the case details (client, role, case type).
- View case analytics, including the number of documents uploaded and pages processed.
- Edit case instructions that every Agent run inside the case will follow.
Deleting a Case
The case admin can permanently delete a case and all of its associated data at any point. This action removes all uploaded documents, processing results, timelines, Agent conversations, and any other data within the case. This action is irreversible, so please ensure you have exported any data you need before proceeding.
Deleting a case is permanent. All documents, processing results, Agent conversations, and timeline data within the case will be permanently removed. Make sure to export any work you need before deleting.
Documents Table
The Documents Table is the entry point for adding or removing data from your case. This is where you upload documents, view the results of Kallam's automatic processing, and interact with your case file in a structured tabular format. While working on a live case, you can continuously upload new documents in bulk and Kallam will process each one individually.
Uploading Documents
To add documents to your case, navigate to the Documents page within the case and upload your files. You can drag and drop files directly or use the upload button. Kallam accepts the following file formats:
- PDF, including scanned PDFs. Kallam applies OCR automatically.
- Word documents (.doc, .docx).
- Images (.jpg, .png, and similar formats).
- Text files (.txt).
- ZIP archives. Kallam extracts the contents and processes each file individually.
The maximum file size for a single upload is 50 megabytes, and the maximum document length is 1,000 pages. Excel documents (.xlsx) are not supported at this time. If you need to include spreadsheet data, convert it to PDF before uploading.
List and Table Views
The Documents Table offers two ways to view your case file. The Table view shows your documents in a structured grid with columns for Title, Date, Category, Folder, Tags, and File, which is ideal for scanning and sorting. The List view gives a more compact reading layout. Switch between them with the view toggle.
Document Details
Click any document row to open its Document Details panel. The panel shows the document's summary, its contextual date, which you can edit, its category, its tags, and your notes. Expanding the panel reveals more structured detail extracted from the document: a longer summary, the parties involved, and the key points. For agreements, it also surfaces the term and the commercial terms. Use this panel to review and correct what Kallam extracted before you rely on it.
Categories
Every document is automatically classified into a category during processing, for example Correspondence, Contract, or Report. Categories are user extendable: if your case needs a classification Kallam does not offer by default, you can add your own and apply it to documents.
Document Splitter
In some cases, your client or counterparty may provide documents as scanned bundles: a single PDF file containing multiple distinct documents that have been scanned together. If you upload a bundled PDF as is, Kallam will treat the entire bundle as one document, which will produce inaccurate processing results.
To handle this, the Documents Table includes a Document Splitter feature. To activate the splitter, click the scissors icon in the Upload Log next to the bundled document. This opens the splitter interface where you can view the document page by page and assign pages to separate subdocuments. Each subdocument is then uploaded and processed separately, ensuring that titles, dates, summaries, and categories are generated correctly for each individual document rather than for the bundle as a whole.
Always split bundled PDFs before uploading them. If you upload a bundled PDF without splitting, Kallam will process it as a single document, and the title, date, and summary will reflect the bundle as a whole rather than the individual documents within it.
Folder Upload
You can upload an entire folder at once, and Kallam recreates its folder structure automatically inside the case. This preserves the organisation you already have rather than flattening everything into a single list.
Document Chronology and Summary Levels
Sorting the Documents Table by contextual date gives you the document chronology: every document arranged in the order events occurred. This is distinct from the event Timeline, which arranges dated facts rather than documents. For each document, Kallam offers three levels of summary so you can choose how much detail you need:
- Brief. A one line summary for quick scanning, ideal when you are reviewing a large number of documents at a glance.
- Comprehensive. A paragraph level summary that captures the main points, useful for understanding the substance of a document without reading the full text.
- Detailed. An in depth summary that preserves key facts, dates, parties, and significant details, for when you need to understand a document thoroughly before reading the original.
Author Details and Activity
Kallam captures author details on the documents you upload, and keeps a log of who on your team added each document and when. This gives the whole team a clear record of how the case file was built.
Retrying Failed Processing
Every document row displays its processing status. If a document fails during OCR or analysis, its status dot turns red and a retry button appears on the document detail drawer. Clicking retry resets the processing state and re-triggers the pipeline for that document specifically, without affecting the rest of the case.
Case Instructions
Next to the view toggle in the Documents page, you will find the Case Instructions drawer. These instructions are free form text that the Agent reads on every turn inside this case. Use them to capture things that are true for the whole case, for example "the Claimant is called ABC Corp", preferred citation styles, or guidance about terminology specific to the dispute.
What to Keep in Mind
- Summarisation and title generation are highly accurate and reliable. These are among Kallam's strongest automated features.
- Contextual date extraction has been significantly improved in the latest update and works well for most document types.
- Categorisation is still undergoing stabilisation. While it works well for common document types (contracts, correspondence, reports), users may occasionally need to correct the category assigned to domain specific or unusual documents.
- Machine written text in English, Arabic, French, German, Spanish, and other languages is processed accurately.
- Handwritten Arabic is less accurately processed than machine written Arabic. The accuracy depends on the clarity of the handwriting.
- Rule of thumb. If the text in a scanned document is easily readable by the untrained human eye, then Kallam's OCR will properly pick it up. However, if the scan quality is poor or the handwriting is not clear, the processing output may contain errors.
- Documents with handwritten data or date stamps. We recommend thoroughly verifying the processing output for any documents that contain handwritten content or dates in stamps, as these may alter the contextual date extracted by the processing pipeline.
Agent
The Agent is Kallam's complete AI assistant, designed for robust document intelligence and information extraction from the documents you have uploaded. Think of the Agent as a junior associate who has thoroughly studied every document in your case file and can provide complete, cited replies to your questions about the case as a whole or about specific documents.
Modes: Reasoning Depth and Output
The Agent has two independent settings you choose before sending a query: how deeply it reasons, and where its output goes.
Reasoning depth:
- Fast. Best for quick, straightforward questions where you need a rapid answer.
- Thinking. Best for thorough or complex queries where accuracy and depth matter. In Thinking mode the Agent reasons more deeply before answering. We recommend Thinking mode for best results.
Output mode:
- Chat. The Agent answers in the conversation thread.
- Compose. The Agent drafts into a document panel beside the chat. See Compose Mode below.
You can combine any reasoning depth with any output mode, for example Thinking together with Compose to draft a careful document.
Querying and Answers
To use the Agent, type your question in the chat interface. After you send a query, the Agent lists the steps it is taking in the Context panel and begins streaming its answer. You do not need to wait for the entire response; the answer appears progressively as the Agent works. Every answer includes source citations that tell you exactly which documents the Agent relied upon.
The Context Panel
As the Agent works, a Context panel on the right shows you exactly what it is doing. It has three parts:
- Progress. The Agent's step by step activity, so you can follow precisely what it did to reach its answer.
- Input Documents. The documents the Agent used as context for this response.
- Generated Artifacts. The work products the Agent produced during the session, such as drafts and extractions. The Context panel is what lets you trust the answer: the reasoning steps, the sources, and the outputs are all in one place.
Source Verification and Citations
Every answer the Agent provides comes with inline source citations, and they distinguish between sources the Agent referenced and sources it analysed in depth. When you click any citation, Kallam opens the original document in a side by side view next to the Agent's reply, scrolled to the exact page where the information was found, so you can verify the answer immediately. Nothing is taken on faith: the Agent shows its work, and you can trace any claim back to its source.
Document Tagging
By default, the Agent searches across all documents in your case. To focus it on one or more specific documents, tag them in your query by typing @ followed by the document name, for example @Supply Agreement March 2022. This tells the Agent to prioritise those documents rather than searching the entire case file. It is useful when you know exactly which document holds the information you need, or when you want the Agent to analyse one document in detail.
Tool Routing and Deep Online Research
The Agent answers from your case documents by default and does not draw on outside information unless you ask it to. It also routes its own tools automatically based on your request, choosing when to extract a table, draft a document, or research online, so you do not have to select a tool yourself.
When your question calls for information beyond the case file, you can ask the Agent to research online, and it will route to Deep online research, a research preview tool. The online sources are listed alongside your case document sources in the answer, so you can verify both. Use online research for context such as legal background, public filings, or industry information; for your primary case work, the Agent stays grounded in your documents.
Compose Mode
Compose is the Agent's output mode for drafting. Selecting it opens a draft document panel beside the chat, and any drafting request you make generates its result in that document rather than in the thread, so you can build a document incrementally through conversation. You can:
- Edit the document directly. Click anywhere in the draft and type, as in a text editor.
- Request edits through chat. Ask the Agent to change specific parts, for example "Expand the section on the payment timeline".
- Select a passage and request changes. Highlight a sentence or paragraph and ask the Agent to modify just that selection.
- Review proposed edits as diffs. When the Agent changes the draft, its edits appear as a diff you can accept or reject, so you stay in control of every change.
- Export the document. Download the draft in Markdown, Word (.docx), or PDF. For best results, use Thinking depth together with Compose.
Agent Settings
You can shape how the Agent works through settings at three levels: organisation, case, and user. The same controls are available at each level: Formality, Detail level, Citation style, and free text Custom instructions. Organisation settings set a firm wide default, case settings apply to everyone working a particular matter, and user settings reflect your own preferences. Case instructions you set here are read by the Agent on every turn inside that case.
Subagents
For specialised work, the Agent draws on subagents: focused workers for tasks such as facts extraction, timeline building, drafting, subchronologies, and search. You do not manage these directly; the Agent calls on them as a task requires, and their output appears as artifacts.
Sessions
Each conversation with the Agent lives inside a session. You can create new sessions, rename them, search through them, and archive the ones you have finished with; archived sessions move out of your active list but remain available. Sessions marked incognito are kept separate from both lists and are intended for quick exploratory queries you do not want to keep. Any session can be exported to PDF, bundling the full conversation and its source citations into a single shareable file.
Tips for Better Results
- Be specific. "What are the key payment milestones in the construction contract dated March 2022?" is much better than "Tell me about payment."
- Reference document types. "In the witness statements..." or "Looking at the email correspondence between X and Y..." helps the Agent focus its search on the most relevant documents.
- Ask follow up questions. The Agent remembers the conversation context. You can say "Now compare that with the opposing party's position" without restating the original question.
- Use for verification. After reviewing a section of your case, ask the Agent to confirm or challenge your understanding with specific citations.
Not Recommended For
- Translation tasks. While the Agent can translate text, it is not optimised for translation and we do not recommend relying on it for accurate translations at this time.
- Nuanced legal analysis. The Agent excels at extracting factual information from documents and discussing specific details it finds. However, it is not designed for nuanced legal reasoning, opinion, or strategic analysis. It is best used as a factual information extraction tool.
- Writing complete drafts from scratch. The Agent is good at rewriting, amending, and building upon existing drafts. However, it should not yet be used to write a complete document from a blank page. Instead, start with an initial template or outline and ask the Agent to modify and expand it based on information from your case documents.
Finder
Finder is Kallam's advanced search tool, allowing you to search for specific words, phrases, or concepts across every document in your case's dataroom. Finder combines traditional keyword matching with semantic similarity, which means it can find not only exact matches but also passages that are conceptually related to your search query.
Match Modes
Finder searches across every document in your case's dataroom, and you choose how it matches:
- Smart. Combines meaning and exact words. This is the recommended default and works for most searches.
- Meaning. Finds passages that are conceptually related to your query, even when they do not contain the exact words. For example, searching for arbitrator also surfaces passages mentioning arbitration, tribunal, and arbitral proceedings.
- Exact. Matches only the words you typed. Ideal for a specific term, a person's name, a date, or a clause number, for example "force majeure". Results from your chosen mode are sorted by relevance, with each match highlighted so you can see the relevant passage in context.
Viewing Results
Each result shows a dynamic snippet of the matching passage. Click any result and Kallam displays the full document on the right, scrolled to the match, so you can read it in full context without leaving the search interface. Click through multiple results to review relevant passages across different documents quickly.
Filters and Saved Searches
You can narrow your search with filters for folder, document type, and date, and you can save a search to reuse it later. Saved searches are useful for queries you return to often as the case file grows.
Use Finder when you know roughly what you are looking for and want to see where specific information lives across your documents. Use the Agent when you want to ask a question and get a synthesised, interpreted answer. They complement each other: Finder shows you where information is, while the Agent interprets and summarises it for you.
Timeline
The Timeline is an event chronology that collects the dated facts found across your case documents and arranges them in the order they occurred. While the document chronology in the Documents Table arranges your documents by date, the Timeline works at the level of individual facts: if a document states "On 15 March 2022, the parties met to discuss the project", that fact appears as its own entry on the Timeline, linked back to its source document. The Timeline is currently a research preview.
How the Timeline Works
The Timeline displays the dated facts extracted from every document in the dataroom. Each entry shows the fact's date, a title, a description, the parties involved, and tags, and it links back to the source document so you can verify it. Because it is generated from the record, the Timeline builds automatically as you upload documents; there is no need to regenerate it.
Working with Facts
You can shape the Timeline:
- Add facts. Add a fact manually. Any fact you add must reference a document already uploaded to the dataroom.
- Delete facts. Remove facts that are irrelevant, duplicated, or inaccurately extracted.
- The Timeline is a research preview and displays all dated facts extracted from your documents. For large case files, this may produce a long list that needs filtering.
- Any fact you add manually must reference a specific document in the dataroom; you cannot add a fact without a source.
Artifacts
The Artifacts page is where every generated work product in a case lives. Anything the Agent drafts, extracts, or compiles for you is saved as an artifact and remains available for review, editing, and export. Tabular extractions produced by the Tabular Extractor also land here, so every piece of AI output in a case is in one place.
Artifact Types
Most artifacts fall into two types:
- Drafts. Documents produced in the Agent's Compose mode, plus any draft you start directly on the Artifacts page.
- Extractions. Structured tables produced by the Tabular Extractor, where each row is a matched document and each column an extracted data point. The Agent can also deliver other kinds of output as artifacts when you ask, for example a report.
Creating and Managing Artifacts
Most artifacts are created as a byproduct of the Agent: ask the Agent to draft something, extract a table, or summarise a set of documents and the result is saved as an artifact automatically. You can also create a draft directly on the Artifacts page. Each artifact supports editing, status (in progress / completed / archived), and export in the formats relevant to that type (Markdown, DOCX, PDF for text; CSV/Excel for tables).
Source Traceability
Artifacts carry their own source citations, mirroring how the Agent works. Every generated fact, quote, or extracted value is tied back to a specific document and page in your case, so you can verify before you rely on it in a submission.
Tabular Extractor (Tabular Extraction)
The Tabular Extractor is the tabular review tool inside the Agent, designed for two main purposes: document identification and information extraction. Think of it as a diligent junior associate who can read through your entire case file and return structured, table based answers to your instructions. No prompt engineering or technical skills required. Each Tabular Extractor query produces an extraction artifact.
The Tabular Extractor is currently available at no additional cost. It may be subject to separate pricing in the future.
Document Identification
The Tabular Extractor can search through all documents in your case's dataroom and identify the ones that match specific criteria you define. For example, you might ask it to find "all correspondences between Party A and Party B sent during 2022" or "all contracts that contain a force majeure clause". The Tabular Extractor will review every document in the case and return only those that match your criteria.
Information Extraction
Beyond just finding documents, the Tabular Extractor can extract specific information from them. The key to using it effectively is to include two things in your query:
- What documents you are looking for. Describe the type of documents you want to find, for example "invoices from 2023" or "emails regarding the delay claim".
- What information you want to extract. Describe the specific data points you want pulled from each document, for example "the amount, the date, and the parties involved".
For example, a complete instruction might be: "Find all invoices from 2023 and extract the invoice number, date, amount, and the parties involved". The Tabular Extractor will identify the matching documents and create a table with each requested data point as a column.
Table Output
The Tabular Extractor returns its results in a structured table format, not as chat style prose. Each row represents a single document that matched your criteria; each column represents a question or data point that the Tabular Extractor has answered about that document. You can export this table to a local file (such as Excel) to continue working on it with your own tools, or share it with colleagues.
Building on Previous Queries
One of the Tabular Extractor's most useful features is the ability to build on a previous query. After receiving your initial results, you can ask it to extract additional information from the same set of identified documents. This new information will appear as new columns added to the existing table, so you can progressively enrich your analysis without starting from scratch.
For example, you might first ask: "Find all contracts and extract the parties and execution date". After reviewing the results, you could follow up with: "Now also extract the governing law and dispute resolution clause from these documents". The Tabular Extractor will add two new columns to your existing table.
Agent vs. Tabular Extractor, when to use which
| Use Case | Agent | Tabular Extractor |
|---|---|---|
| Ask a specific question about the case | Best choice: conversational answers with citations | Can do it, but returns tables instead of prose |
| Extract structured data from documents | Limited: returns prose, not tables | Best choice: returns structured, exportable tables |
| Compare information across documents | Good for narrative comparisons | Best for side by side tabular comparisons |
| Draft a document or submission | Best choice: use Compose mode | Not designed for drafting |
- After submitting a Tabular Extractor query, you may need to refresh the page for the results to appear. If you do not see the extraction populate, try refreshing.
User Profile
Your User Profile contains your personal account settings and security configuration. These settings apply to your individual Kallam account across all cases and organisations you belong to.
Password Management
You can change your Kallam account password at any time from your User Profile settings. We recommend using a strong, unique password that you do not reuse across other services.
Multi-Factor Authentication (MFA)
For enhanced security, you can enable multi-factor authentication on your account. MFA adds an extra layer of protection by requiring a second verification step when you log in. You can link your Kallam account to your organisation's MFA application (such as Microsoft Authenticator, Google Authenticator, or similar), ensuring that even if your password is compromised, your account remains secure.
Single Sign-On (SSO) Integration
For convenience, you can link your Kallam account directly to your Google or Microsoft account. This allows you to log into Kallam using your existing Google or Microsoft credentials, eliminating the need to remember a separate password. This is especially useful for firms that already use Google Workspace or Microsoft 365 for their daily operations.
If you are locked out of your account, contact your organisation admin or reach out to info@kallam.ai for account recovery assistance.
Word Add-In
The Kallam Word Add-In allows you to manage citations in your briefs automatically, completely automating the exhibit numbering task as well as the generation of the list of exhibits. It brings your Kallam case intelligence directly into Microsoft Word, so you never need to leave your document to reference your case file.
Setup and Connection
To get started with the Word Add-In, open Microsoft Word and install the Kallam Add-In from the Add-Ins panel. Once installed, the Add-In appears as a panel on the right side of your Word document. Sign into the Add-In using your existing Kallam credentials, then select the case you are working on. The Add-In will connect to your case's document library, making all uploaded documents available for citation.
Configuring Citation Prefixes
Before inserting citations, configure the prefixes that will be used for your exhibit numbering. The Add-In distinguishes between Factual Exhibits and Legal Exhibits, and you can set a different prefix for each. For example, you might use "C" for Claimant factual exhibits and "CL" for Claimant legal exhibits, or "R" and "RL" for Respondent exhibits. These prefixes will be used when exhibit IDs are assigned later.
Inserting Citations
To cite a document in your brief, follow these steps:
Position Your Cursor
Place your cursor at the exact location in your brief where you want the citation to appear.
Click Insert Citation
In the Kallam Add-In panel, click the Insert Citation button. A search dialog will appear.
Search for and Select the Document
Search for the document you want to cite by typing its name or keywords. Only documents that have been uploaded to the case can be cited. You cannot cite external documents that are not in Kallam.
Select Exhibit Type
Indicate whether this is a Factual Exhibit or a Legal Exhibit. This classification determines which prefix will be used when exhibit IDs are assigned.
When the citation is inserted, two things happen:
- A footnote is placed at the position of your cursor in the document.
- A placeholder appears in the footer of the page, showing the date of the cited document and its name.
List of Exhibits
As you insert citations throughout your brief, the Word Add-In automatically maintains a List of Exhibits at the end of the document. This list includes every document you have cited, making it easy to generate the exhibit list that is typically required as part of arbitration or litigation submissions.
Assigning Exhibit IDs
Once you have finished inserting all citations in your brief, click the Assign Exhibit IDs button in the Add-In. The tool will then:
- Automatically number all citations in the footer based on their order of appearance in the document.
- Distinguish between Factual Exhibits and Legal Exhibits, numbering them separately using the prefixes you configured earlier, for example C-1, C-2, C-3 for factual and CL-1, CL-2 for legal.
- Update the List of Exhibits at the end of the document with the assigned IDs.
This automation eliminates the tedious and error prone task of manually numbering exhibits, which traditionally requires careful tracking and frequent renumbering as the brief evolves.
- The placement of the date in the footer follows a standard format. If you wish to change the style or layout of the footer, for example to match a specific tribunal's formatting requirements, you will need to make those adjustments manually in Word for now.
The Word Add-In works with both the desktop version of Microsoft Word and Word for the web. For the best experience, we recommend using the desktop version.
Your Data. Your Control. Always.
Kallam is built from the ground up to meet the confidentiality requirements of international disputes work. Your case documents are among the most sensitive materials in legal practice, and we treat their security accordingly.
Azure EU Hosting
Hosted on Microsoft Azure in the EU, inside the European Union.
Never Used to Train AI
Your documents and prompts are never used to train AI models, and nothing sent to the models is retained after processing. Your data stays encrypted and under your firm's control.
Encryption and MFA
Every document is encrypted with AES-256 at rest and over TLS 1.2 or above in transit. Multi-factor authentication is enabled by default.
Full Client Data Isolation
Complete data isolation between clients and cases. Designed for the confidentiality requirements of legal practice.