Logo
Home

Ask Ark: AI-Powered Learning Assistant

Wall Street Prep: "Ask Ark" the AI-Powered Learning Assistant

An AI assistant combining LLM capabilities with a RAG pipeline to deliver real-time, accurate answers — transforming how finance professionals engage with educational content.

INDUSTRY

FinTech/EdTech

SERVICE

AI Application Engineering

DELIVERY MODEL

Two-Phase Build

SUMMARY

A smarter way to navigate financial education

"By combining LLM capabilities with a RAG pipeline, Ask Ark delivers real-time, accurate answers — creating a smarter, more responsive learning experience."

Wall Street Prep, a leading financial training platform, partnered with Waverley to develop Ask Ark, an AI-powered assistant designed to transform user engagement with financial education content.

By combining Large Language Model (LLM) capabilities with a Retrieval-Augmented Generation (RAG) pipeline, Ask Ark delivers real-time, accurate answers to learner queries and directs them to the most relevant platform content.

This creates a smarter, more responsive learning experience that aligns with Wall Street Prep's mission to provide world-class financial training.

ABOUT THE CLIENT

Wall Street Prep

Wall Street Prep is a well-established provider of financial training, offering a comprehensive content library that includes articles, video lessons, and audio materials.

The platform serves a highly motivated audience of finance professionals, students, and analysts who require precision, speed, and depth in their resources.

As the platform's content library grew, so did the challenge of helping users navigate it efficiently. Learners needed a way to get immediate, accurate answers to their questions without having to manually search through hours of video or dozens of articles.

Wall Street Prep recognized an opportunity to leverage AI to bridge that gap — not just as a search tool, but as an intelligent learning companion embedded directly within the platform.

Project Analysis & Challenges

The core challenge: discoverability and immediacy

Despite having a comprehensive and high-quality content library, Wall Street Prep faced a core user-experience challenge. Learners often struggled to find the exact answers or content sections they needed in the moment, leading to friction in the learning journey and potential disengagement.

Content Navigation at Scale

With a large and growing library of articles, videos, and audio, users struggled to pinpoint the most relevant content for their specific questions.

Lack of Real-Time Support

There was no mechanism for users to receive instant, intelligent responses to their queries without waiting for instructor feedback.

Measuring Learning Engagement

The platform needed clearer, data-driven signals — such as query volume and satisfaction scores — to understand how users were interacting with content.

Accuracy Requirements

In the financial domain, incorrect or imprecise answers have significant consequences. Any AI solution would need to meet high standards for factual accuracy, validated by financial experts before deployment.

THE SOLUTION

Built in two phases, delivered with precision

Waverley's global team — spanning Europe and Latin America — worked with Wall Street Prep to develop Ask Ark. The solution was built in two phases: an MVP focused on core AI functionality, followed by an expansion phase introducing more advanced capabilities.

PHASE 1

MVP — Core AI Assistant

The MVP established the foundational infrastructure for Ask Ark, focusing on three key components:

  1. Natural Language Processing (NLP) and Query Understanding

The system uses NLP to parse and understand user intent, extracting key information from queries to ensure responses are contextually relevant and precise.

  1. RAG Pipeline and Knowledge Base

A continuous data pipeline was built to extract text content from Wall Street Prep's CMS, pre-process it, and index it for efficient retrieval. This indexed knowledge base powers the RAG pipeline, grounding every AI response in authoritative platform content rather than relying solely on general model knowledge.

  1. OpenAI GPT-4 API Integration and Response Generation

The core AI engine integrates with the OpenAI API to generate concise, informative answers. A standout feature is deep content linking: the system directs users to the exact section of an article, video timestamp, or audio transcript marker most relevant to their query. Deployed on AWS S3 and CloudFront with a React-based chat interface embedded natively within the platform.

PHASE 2

Expansion — Advanced Capabilities

Building on the MVP, the expansion phase introduced:

  1. Personalized Recommendations

Analyzing user interactions and learning preferences to surface tailored course and resource suggestions.

  1. Contextual Understanding

Enabling the assistant to handle multi-part questions and follow-up inquiries seamlessly, incorporating course-specific context via an LMS API integration.

  1. Quality Evaluation System

An LLM judge model was implemented to score Ask Ark's responses against instructor-provided answers, with sentiment analysis planned to further gauge user satisfaction.

Tech Stack at a Glance

AI/ML
BACKEND/DATA
FRONTEND
INTEGRATIONS/PROCESS
OpenAI GPT-4 API
CMS
React
OpenAI API
RESULTS & OUTCOMES

Measurable improvements across engagement and efficiency

Real-Time Query Resolution

Users now receive instant, accurate answers to financial training questions directly within the platform, eliminating wait time previously associated with instructor-led support.

Deeper Content Engagement

By linking users to specific content sections — article paragraphs, video timestamps, audio markers — the assistant drives more purposeful and targeted interaction with the content library.

Improved Internal Efficiency

Refactored data pipelines enabled faster content processing and transactional updates, accelerating the team's ability to implement new features.

Accuracy Validation Framework

A rigorous evaluation system — including an LLM judge model benchmarking responses against instructor answers — ensures the assistant maintains high accuracy standards required in financial education.

Scalable Measurement Infrastructure

User engagement metrics — queries handled, satisfaction ratings — are now actively tracked, providing Wall Street Prep with actionable data to continuously optimize the learning experience.

Quote Icon.png

The Ask Ark project demonstrates how a well-architected AI assistant — built on RAG, NLP, and LLM technology — can elevate the user experience on a specialized educational platform.

Rather than replacing the expertise of Wall Street Prep's instructors, Ask Ark amplifies it by making their knowledge instantly accessible, precisely navigable, and continuously improving through data-driven evaluation.

For platforms with extensive content libraries and knowledgeable audiences, AI is not just a convenience feature — it is a strategic lever for engagement, retention, and learning outcomes.

Let's build your AI-powered experience.

Is your platform rich in content and data that could be leveraged? We can help you build an AI solution that improves customer experience, maximizes your data's potential, and advances your company to the next level.