HL V1 Mediation App Final Outline #2

Overview

What is Elerion?

Elerion is an AI powered virtual mediation service. Leveraging the power of chatGPT, Speech To Text, and Text To Speech to resolve mediation disputes between two parties.

Version One of the Elerion Mediation App is focused on achieving two primary goals:

  • Break functionality down into components that can be re-used for other business ventures
  • Built in-house using time and resources as efficiently as possible to reduce cost

Both goals can be accomplished by building the product in three core, separate, components

  1. Real time, audio based, legal problem solving AI network Through creative and strategic problem solving we’ve successfully outlined a network of AI’s that can be used for more than just Mediation but rather universal legal problem solving done in real time audio. Creating the AI network separate from the other two components will allow it to easily be added onto Zoom, Virtual Phone API’s and more. Applications of this range from initial consults, lead intake, mediation and more. Further on in this document we’ve outlined how this will be achieved.
  2. User registration and management Building this separate from the other two components will allow for it to be re-used in services like MWLP, Tenant Defense and any other app thought of as user registration and management is a universally needed functionality. We will tie Wordpress, Stripe, and the Airtable CRM together to allow for a foundation that can be easily extended with wp plugins.
  3. Virtual Meeting Rooms Creating stand-alone, simplified, Zoom functionality managed in-house will allow for easy application throughout the HepworthLegal law firm.

Estimated Time To Complete

The estimated time required to build the app is: min 457 hours, max 577 hours

  1. Initial testing and scoping of technical theory, project management outlining: 50 hours
  2. Real time, audio based, legal problem solving AI network: min 250 hours, max 340 hours
  3. User registration and management: 67 hours
  4. Virtual based meeting room’s: min 60 hours, max 90 hours
  5. Front Facing Website and Copy: 30 hours

Core Components Functionality Breakdown

Component 1: Real time, audio based, legal problem solving AI network

Estimated time to complete

Still working to outline final task list

Features to be created

  • Network of AI’s fine-tuned and tested to solve legal problems
  • Ability to append network to third party software
  • Information and conversation database architecture that weights information given as it’s stored for related to the listed problem and desired goal
  • Semantic search architecture that pulls information necessary to achieve desired stages outcome
  • Seamlessly adjust and switch LLM used for each stage
  • Quickly adjust prompt for each stage

Originally the mediation process was outlined in specific stages AI could follow to get to a resolution. But with clever abstraction it was realized that the process could be set in a way that allowed for an overarching structure that could solve all legal problems efficiently, regardless of matter.

Legal Problem Solving AI Network Cycle

Overview: Each stage will contain a fine-tuned and tested prompt that aims to achieve just the desired goal of the stage. With set conditions for failure that repeat back into previous stages it allows for a simple, continuous cycle to solve any legal problem.

Stages Detailed Outline

Identify Problem & Define Goal Stage 1

Using pre-meeting information provided and initial communication with client clearly define problem and the desired goal. Get clear confirmation from client that problem and goal are correctly defined, this will be used as AI’s north star throughout the Problem Solving Cycle

Move to the next stage once confirmation is given of correctly defined problem and goal.

Generate Questions Stage 2

Using the problem, desired goal, and any information provided to this step generate information needed to achieve goal. From the information needed generate a list of questions to ask the client. Weight the questions in terms of important to generate a solution to achieve the goal.

Move to the next stage once questions are generated.

Gather Information Stage 3

Ask all questions generated, weight and store the information gathered.

Move to the next stage once all questions are answered

Generate Solutions Stage 4

Use information gathered and generate at least 2-3 possible solutions that achieve the outlined goal.

Move to the next stage once all questions are answered.

Solution Evaluation Stage 5

Do an internal review of the generated solutions, weighting them from best to worst and if they cross a minimum threshold to be a presentable solution.

If no solutions pass the minimum threshold then pass to Adjust Prompt stage which determines why the solutions are bad.

  • If more information is needed then adjust prompt and go back to Stage 1.
  • If solution doesn’t meet desired goal then list why in prompt and go back to Stage 4.

Move to the next stage if any solutions pass the minimum threshold

Present To Client Step 6

If client likes the solution, then cycle is completed, problem solved.

If client does not like the solution then ask ask why.

  • If more information is needed then adjust prompt and go back to Stage 1.
  • If solution doesn’t meet desired goal then list why in prompt and go back to Stage 4.
Double click to see flowchart
Double click to see flowchart

Component 2: User registration and management:

Estimated time to complete

67 hours

Description

A secure portal will allow signed-up users to log in and access additional features, including the ability to add more users for the meeting, schedule a meeting time with the other primary user, add pre-mediation notes, obtain meeting access links, pay for mediation after it is completed, and view draft agreements. User management, follow-up, and the database will be easily controlled by Hepworth because it will be built around Airtable.

Features to be created

  • Payments
  • Airtable CRM to manage users and database
  • Portal on frontend website connected to Airtable for user management
  • User management, primary users that can add basic users
  • Add and update user specific information
  • Schedule mediation time between two primary users
  • Notifications on actions and upcoming meeting using existing Airtable CRM functionality
‣
Click to see detailed task-list with hours breakdown

Component 3: Virtual Meeting Rooms

Estimated time to complete

Min: 60 hours

Max: 90 hours

Description

Creating a stand-alone, simplified, Zoom functionality managed in-house will allow for easy application throughout the HepworthLegal law firm and an administrator interface that will run on AirTable.

Features to be created

  • By realizing the data exchange with LLM via API, our frontend and backend will be separated from each other.
  • Meetings will be password-protected, participants will be identified and the duration will be limited.
  • It will be able to grow horizontally through structures such as load balancer with API infrastructure.
  • When the mediation meeting is completed, all conversations will be uploaded to AirTable in PDF format.
‣
Click to see detailed task-list with hours breakdown

Additional Component: Front Facing Website and Copy

Utilizing OWC’s streamlined website process a front facing homepage, get started, pricing, and about page can be developed quickly and efficiently

Estimated time to complete

30 hours

Features to be created

  • Persuasive copy that clearly defines the product and it’s unique value to the customer
  • The following pages fully designed
    • Home
    • Get Started
    • Pricing
    • About

Path To Build Elerion

Problem

The journey of creating the AI Legal Problem Solving Network, particularly with its ambitious goal of functioning as a Mediator, is fraught with complexities. The intricacies of managing multi-party conversations lasting 5-8 hours magnify testing and refining challenges. While we strive for unparalleled problem-solving and memory capacities from the outset, this ambition inadvertently extends the launch timeline, shrouding it in uncertainty. As the project unfolds in-house, a tangible pressure mounts by the 2nd or 3rd month — a pressure for a launch to justify the investments, even if the product isn't optimally ready. Launching prematurely might expose users, unfamiliar with AI-led legal audio interactions, to sessions exceeding 8 hours, presenting unforeseen user-experience challenges. This intersection of indeterminate development timelines and the absence of user feedback amplifies the risks associated with the debut version's intricate functionalities.

Solution

To mitigate these challenges, a phased development approach is proposed: Phase 1 focuses on "Elerion For Divorce," followed by Phase 2, "Elerion For Mediation." This strategy fosters a balance between innovation and pragmatic roll-out, ensuring quality and user receptiveness.

Phase 1: Elerion For Divorce

Rather than diving straight into full-scale mediation, a strategically prudent step would be to introduce Elerion as a tool for divorce-related consultations. This phase will allow users to seek guidance from Elerion about their issues, culminating in the provision of pertinent documents and advice.

Benefits of this phase include:

  • Simplification: Divorce consultations allow for a more controlled environment. It becomes easier to program the AI to handle the nuances of these discussions.
  • Foundation: The lessons learned and the technological underpinnings of this phase will serve as the base for the mediation tool.
  • Feedback: Engaging users early will provide insights to refine Elerion for subsequent phases.
  • Economic Prudence: Faster deployment and fewer developmental complexities imply less capital outlay before ROI is realized.

Broader Applications

The potential of Elerion's technology extends far beyond the domain of divorce settlements. Its AI-driven capabilities can reshape the legal landscape and offer versatility in various business endeavors:

  • New Practice Areas: Elerion can be adapted for various legal fields, such as tenant defense and contract law, offering tailored guidance across multiple specialties.
  • Front-End Firm Assistance: Elerion can manage incoming calls to law firms, determining the client's needs and directing them appropriately.
  • Consultations: The platform can replace attorneys during initial consultations, providing clients with immediate insights and legal perspectives before a detailed human intervention.

Estimated Time and Cost

Feature
Time
Cost
Initial Testing, Technical Theory, & Project Management
40 hours
$4,000
Real time, audio based, legal problem solving AI network
175-265 hours
$11,000 - $16,560
User Registration and Management
30 hours
$3,000
Virtual Meeting Rooms
55-85 hours
$4,125 - $6,375
Front Facing Website and Copy
30 hours
$3,000
Total
330-450 hours
$25,125 - $32,935

Phase 2: Elerion For Mediation

After the initial deployment and feedback collection, Phase 2 will escalate the functionality to include mediation. The primary challenges here will involve equipping Elerion for prolonged interactions and facilitating discussions between multiple parties.

Feature
Time
Cost
Initial Testing, Technical Theory, & Project Management
20 hours
$2,000
Real time, audio based, legal problem solving AI network
75 hours
$7,500
User Registration and Management
37 hours
$3,700
Virtual Meeting Rooms
5 hours
$500
Total
137 hours
$13,700

Next Steps

  1. Confirmation of funds to begin building Elerion
  2. Begin Initial Testing, Technical Theory, & Project Management creating an estimated time to launch to keep the project on track
  3. Begin development once timeline approved