GoalSetter (Students of Wealth, Inc.)
New York, NY
Financial Education / EdTech
Amazon Bedrock AWS (Non-Production Environment) RAG (Retrieval-Augmented Generation) Vector Database Large Language Models (LLMs) JSON-Structured Content Schemas
GoalSetter, the Students of Wealth, Inc., is a financial education platform dedicated to teaching money management skills to young learners from middle school through college. Facing the challenge of delivering personalized, age-appropriate financial lessons at scale, GoalSetter partnered with Avahi to build an adaptive learning system powered by Generative AI and Retrieval-Augmented Generation (RAG) on AWS. Avahi designed and developed a deterministic, syllabus-grounded content engine that classifies students by knowledge tier, personalizes lesson content to each learner’s interests and goals, and adapts tone and style based on age, all without sacrificing accuracy or consistency. The result is a scalable, auditable AI learning platform that brings GoalSetter one step closer to delivering truly individualized financial education at a national scale.
GoalSetter, operating as Students of Wealth, Inc., is a New York-based financial education company on a mission to close the wealth gap by teaching financial literacy to young Americans. Their platform serves students across middle school, high school, and college, offering structured curriculum, quizzes, and assessments designed to build lasting money skills. Operating in the intersection of EdTech and financial education, a regulated-adjacent domain serving minors, GoalSetter requires an unusually high bar for content accuracy, pedagogical integrity, and age-appropriate delivery.
GoalSetter’s existing system delivered financial education content in a largely static, one-size-fits-all format. Educators were expected to manually adapt lessons to suit students at different knowledge levels, with different learning styles, interests, and age groups. This approach was neither scalable nor consistent, and it placed an unsustainable burden on the pedagogy team as GoalSetter’s student base grew.
The deeper challenge was personalization at scale. A middle schooler just learning about saving behaves very differently as a learner than a college student exploring investment strategies. Delivering the right content, in the right tone, tied to the right personal context, required a level of dynamic content generation that GoalSetter’s existing infrastructure simply couldn’t support.
Without a solution, GoalSetter risked delivering generic lessons that failed to engage students, weakened learning outcomes, and forced educators into endless manual content adaptation cycles. In a domain where trust, accuracy, and engagement are everything, especially when the audience is young learners, the cost of inaction was high. Scalability would remain out of reach, and GoalSetter’s mission to deliver financial literacy broadly would be limited by the bandwidth of its content team.
AWS provided the cloud infrastructure and AI/ML services necessary to build a production-grade, secure adaptive learning system. Given the sensitivity of the domain — financial education content for minors — the non-production AWS environment offered by Avahi gave GoalSetter a safe, isolated space to validate the system’s capabilities without exposing any live systems or production data. AWS’s ecosystem, including Amazon Bedrock for foundation model access and purpose-built services for vector storage and retrieval, made it the natural platform for deploying a RAG-based GenAI solution with the scalability, compliance posture, and reliability GoalSetter required.
Avahi brought a rare combination of AWS technical depth and applied GenAI expertise to this engagement — delivered as an AWS Innovation Waves (IW) Build, a structured, accelerated solution development program available to select AWS partners. This program gave GoalSetter access to a fast-tracked, milestone-driven build with clear accountability, expert guidance, and AWS-backed credibility.
Beyond the program structure, Avahi’s specific experience architecting deterministic GenAI pipelines for high-stakes domains made them the right fit. GoalSetter needed a partner who understood that in EdTech — particularly for a regulated-adjacent context serving minors — AI outputs couldn’t just be “good enough.” They needed to be grounded, consistent, and auditable. Avahi’s approach to RAG architecture, metadata-tagged embeddings, and structured educator review checkpoints gave GoalSetter confidence that the system could meet both pedagogical and operational standards.
Avahi designed and built an end-to-end adaptive learning system that uses Generative AI and RAG to deliver personalized financial literacy lessons grounded in GoalSetter’s official syllabus. The architecture centers on two foundational questionnaires: a Placement Questionnaire that classifies each student into a knowledge tier, and a Profile Questionnaire that captures personal context: interests, goals, and learning preferences. Together, these inputs feed a RAG pipeline that retrieves syllabus-aligned content and generates lessons tailored to each individual learner.
At the core of the solution is a deterministic RAG pipeline built on paragraph-level, metadata-tagged embeddings of GoalSetter’s content library. Rather than relying on open-ended LLM generation, every lesson output is grounded in retrieved source material from GoalSetter’s approved curriculum. Low temperature settings and deterministic query design ensure that the system produces consistent, reproducible outputs, a critical requirement in an educational context where content drift and hallucination pose real risks to student outcomes and platform trust.
The lesson personalization engine adapts not only the content of each lesson but also its tone and reading style based on the student’s age group — differentiated across middle school, high school, and college learners. Each generated lesson includes a personalized anchor tied to the student’s stated goals or interests, making financial concepts more relevant and engaging. Matched quiz pairs are generated alongside each lesson to reinforce learning and support assessment continuity.
AWS services central to the solution include Amazon Bedrock for foundation model access and inference, a vector database for lesson embeddings and semantic retrieval, and a structured JSON content schema for lesson ingestion and metadata management, all running within a secure, non-production AWS environment provisioned and managed by Avahi.
The system was validated against GoalSetter’s syllabus through structured educator review checkpoints, shifting the content team’s role from manual review of every output to focused validation at defined milestones. This significantly reduced review overhead while maintaining GoalSetter’s pedagogical standards across all generated content.
readiness to move from manual content adaptation to AI-powered personalized education at scale. The system successfully classifies students into learning tiers, adapts tone by age group, and generates personalized, syllabus-grounded lessons and quizzes — all with deterministic consistency. Educator validation effort was significantly reduced by shifting reviews from every individual AI output to structured checkpoint reviews, freeing the content team to focus on curriculum strategy rather than content QA.
The solution also establishes a replicable architecture that can be extended beyond the initial lesson set as GoalSetter scales. With the adaptive learning engine validated, GoalSetter now has a clear pathway to full content ingestion and production deployment, bringing personalized financial education within reach for a much larger student population.
Target Performance Metrics Validated Against:
Let’s explore your high-impact AI opportunities together in a complimentary session
