Smart Assistant
Introduction
Odcanit Platinum is Israel’s leading legal CRM, with over 4000 law firms.
Lawyers deal with a huge amount of information. Every case has dozens of folders and hundreds of documents, making it a nightmare to find a specific detail, summarize what's going on in the case, or create a new document from the case data.
The goal of this project was to simplify this mess. We built an AI Chat that lives inside the system. Instead of digging through files, users can just ask the chat questions about their case. But we didn't stop at just talking; we made the chat functional so users can create meetings and tasks directly from the conversation.
My Role
Conducting interviews and observations to understand user pain points and analyzing usage data
Designing the conversational flow, defining how the AI interacts with users and how it "thinks" and responds to complex legal queries.
Creating a clean, modern interface and building the interactive components
Project Cope
Team - Myself, Product Team lead, Development, and Product team.
Duration - 7 weeks from scratch to production
Tools
Design - Figma
Collaboration - Interior systems
The Problem
Users spend lots of time on case actions, document creation, and studying case files
Users spend a significant amount of time manually studying case files to find critical information, which slows down the entire legal process. On top of that, creating documents from scratch and manually updating tasks or meetings makes the whole process inefficient and time-consuming.
4 Min'
In average for a user to find a specific detail in a document
61
The average number of documents per case for a user
4 Hours
Manually reviewing and summarizing information
3 Hours
Spends daily on performing actions and data entry
As one of our users said…
I'm looking for an AI that functions like an intern: drafting legal documents, performing deep case analysis, and summarizing archives instantly.
A partner that stays with the firm forever and never stops learning
Eran, Lawyer
The Solution
Create a smart assistant that understands the case files, drafting documents, and handles manual operations to let users focus on strategy, not paperwork
THE PROCESS
Discovery & User Research
To understand the friction in the current legal workflow, I conducted a deep dive into user behavior through 3 main channels:
User Survyes
Answered by a wide range of users to identify the most time consuming daily tasks, their needs and how can we help
User Interviews
Conducted interviews with dozens of users to understand their pain points, and how they are working today with external AI solutions
Observation
Observed how users currently interact in their daily tasks and with AI tools, and where the workflow breaks
Identifying the Pain Points
From the research, I identified the key pain points that cause the most frustration and slow down the daily
legal workflow:
Search Struggle
Too much time is wasted searching for specific data across endless files, taking focus away from actual legal strategy
Manual Recap Effort
When a user opens a case after a few weeks, there’s no quick way to get a 'briefing' on the latest updates or key issues without digging through everything manually.
Drafting documents
Starting every document from a blank page or an old template, and reading dozens of documents from the case, is slow and leads to repetitive manual work.
Disconnected Workflow
To perform a simple task like scheduling a hearing, or record an expense, users have to leave their legal professional work and navigate through multiple screens.
User Goals
To ensure the solution is effective, I defined three primary goals for the user:
Ideation & Concept
The core idea was to shift from "searching and clicking" to "asking and doing."
Ask, don't search
Stop digging through folders. Just ask the AI where the info is.
Zero background needed
The AI is already 'in' the case file. No need to explain the context.
Answers to Actions
For example: when the AI finds a date, turn it into a calendar event in one click.
Instant case briefing
Summarize 200 pages into 5 bullet points. Catch up in seconds.
Easy drafting
Let the AI write the first draft based on the case facts.
Just review and send
Talk to the Data
Natural language only. No complex filters or technical searches
The Final Solution
A context-aware AI assistant that turns hours of document digging into seconds of productive work
Dynamic Document Interaction
Users can upload documents and let the assistant summarize key points, prepare for hearings and meetings, or generate new legal drafts based on their specific case files.
Case Insights & summary
The interface enables a holistic conversation with the entire case file.
Users can extract complex financial data, generate comprehensive case summaries for client meetings, and receive strategic recommendations for next steps, all based on analysis of the case data.
From Text to Action
Users can convert insights into immediate administrative actions, like scheduling meetings, sending automated email invites, and creating task lists or reminders.
It bridges the gap between legal analysis and office management in one seamless flow.
Impact & Feedback
The goal was to transform the AI from a "nice to have" feature into a reliable team member.
We measured success by how much manual work was eliminated and how confident users felt in the AI’s results. The data we collected from users showed that the solution became a core part of their workflow almost immediately:
54%
Used the solution at least 5 times in one montn
84%
Less time spent searching through case files
3X
Faster drafting of initial legal documents
47%
Of AI answers converted directly into actions
Hear From the Users
What Do I Learned?
I learned how to balance user control with AI automation. The challenge wasn't just making the AI smart, but making sure the user still feels in the driver’s seat. It’s about creating a partnership between the human and the machine, not replacing one with the other.
Explore More Projects
User Onboarding
Odcanit LTD. • 2024
Accelerating users and company's time by replacing manual set-up by with an intuitive onboarding flow
View Case Study
Platform's Redesign
Odcanit LTD. • 2025