LitifyAI Drafting

LitifyAI Drafting

An AI-generative drafting assistant that captures a firm’s unique voice to produce legal documents in minutes.
DETAILS

Attorneys spend a significant portion of their day on routine drafting, but generic AI outputs often fail to meet the rigorous standards of legal formatting and tone.

My goal was to move beyond a standard chatbot interface to create a context-aware drafting engine. I led the design of "Blueprints," a system that allows AI to adopt a firm's specific "voice" and formatting standards, reducing draft time, while ensuring 100% compliance with jurisdictional requirements.

ROLE

Senior Product Designer, AI and Core Practice Management | Litify | 2025

Strategic Design Vision and Roadmapping, AI-Driven Design, Stakeholder Alignment, End to End Design, Prototyping, User Research and Testing

The Challenge

We discovered that while law firms wanted the efficiency of AI, strict data privacy and the fear of generic output were major adoption barriers. Additionally, every law firm has a unique stylistic "fingerprint." A generic AI draft actually increases work if the attorney has to rewrite it to match their style. I had to design a way for the AI to "learn" from a firm's existing document library without compromising data integrity.

Strategic Insights

My research revealed a critical misalignment in the initial product assumptions: drafting the text wasn't the primary bottleneck, re-formatting was. Legal documents must adhere to rigid jurisdictional and firm-specific standards, that vary by case type and document type. We discovered that if the AI’s output required manual adjustment of formatting, we would face a near-zero adoption rate. Paralegals and attorneys simply won't use a tool that creates 'extra work' in the final mile of document preparation. I shifted the focus from a standard 'Chat' interface to a 'Template-first' architecture. This led to the creation of Blueprints, where users select existing 'gold-standard' documents to serve as formatting guides. By ensuring the AI adopts the firm's specific 'visual voice' from the start, we transformed a potential adoption failure into a seamless, court-ready drafting engine.

Defining the MVP

Defining the MVP

While the long-term goal was to build a total 'Intelligence Ecosystem,' I stepped in to lead a prioritization exercise to make sure we launched a stable, high-value product rather than getting bogged down in over-engineering. We mapped our ideas onto a Value vs. Complexity matrix, which led to the tough call to pull a Dynamic Prompt Library from the initial MVP.

Even though having a central repository for high-performing prompts is a great way to turn the tool into a governed asset, our research made it clear that solving the 'formatting roadblock' with Blueprints was the do-or-die requirement for adoption. By deferring the library, we cut down on engineering complexity and put all our focus into the Source-Grounded Accuracy features that were non-negotiable for building user trust. This move allowed us to hit our launch date with a 'trust-first' product that solved the most painful friction point on day one.

While the long-term goal was to build a total 'Intelligence Ecosystem,' I stepped in to lead a prioritization exercise to make sure we launched a stable, high-value product rather than getting bogged down in over-engineering. We mapped our ideas onto a Value vs. Complexity matrix, which led to the tough call to pull a Dynamic Prompt Library from the initial MVP.

Even though having a central repository for high-performing prompts is a great way to turn the tool into a governed asset, our research made it clear that solving the 'formatting roadblock' with Blueprints was the do-or-die requirement for adoption. By deferring the library, we cut down on engineering complexity and put all our focus into the Source-Grounded Accuracy features that were non-negotiable for building user trust. This move allowed us to hit our launch date with a 'trust-first' product that solved the most painful friction point on day one.

The Solution

Instead of a generic AI text generator, I architected an intelligent ecosystem that uses a firm’s gold-standard work as its foundation. The core is a framework we called "Blueprints," which acts as a bridge between AI and the strict stylistic demands of legal work by using existing documents as formatting guides. I also designed a Source-Grounded Accuracy UI that lets attorneys instantly audit the AI's logic against original case data. By solving for the final mile of formatting and trust, we transformed a potential liability into a high-fidelity, court-ready asset

drafting-blueprints
drafting-blueprints
drafting-blueprints

Blueprints

I architected the "Blueprints" feature to allow users to select gold-standard documents from past cases to serve as structural guides . This ensures that the AI-generated output matches the firm’s specific margins, headers, and citation styles from the start, bypassing the formatting hurdles that typically lead to product abandonment .

drafting-review-sources
drafting-review-sources
drafting-review-sources

Source-Grounded Accuracy

To address the industry's Zero-Trust requirement, I designed a verification UI that explicitly links generated text to its underlying case data . This allows attorneys to audit the AI's logic instantly, ensuring that names, dates, and figures are accurate and reliable before the document is finalized

"Emily’s product knowledge, technical expertise, and UX expertise make her a triple threat. She’s essential in shaping our North Star vision for how Litify’s AI products come together as an ecosystem."

Liz

Director of AI Product and Commercialization @ Litify

Ready to Work Together?

Ready to Work Together?

Ready to Work Together?