Case study · Financial advisory (RIA)
Plan Assist.
A production-ready AI-powered financial advisory platform built for a boutique Registered Investment Advisor (RIA) firm.
- What we built
- Compliance-first AI advisory web platform
- Stack
- Next.js/React framework, Responsive design (mobile and desktop), Clean, Google-inspired UI, Real-time chat interface
- Services
- Web appAI assistantAdvisor dashboard
About the project
What is Plan Assist?
This project involved the design and development of a production-ready AI-powered financial advisory platform built for a boutique Registered Investment Advisor (RIA) firm serving approximately 50 household clients. The goal was to create a secure, scalable system that enhances advisor-client communication through AI while strictly adhering to SEC and FINRA regulatory requirements. Unlike typical AI chat or fintech platforms, this solution was architected around a three-way communication model: Client, Advisor, AI Assistant. The AI assistant supports education, clarification, and structured summaries, but never replaces the advisor. Regulatory compliance, auditability, and human oversight are embedded into the platform at every level.
Built for real client usage
The platform was designed for real client usage from day one, not as an MVP or demo. It emphasizes production readiness, security, data privacy, and compliance-first AI behaviour, while maintaining a clean, minimal, and intuitive user experience inspired by Google's design philosophy.
Conclusion
This AI-powered financial advisory platform demonstrates how AI can be safely and effectively integrated into regulated industries when designed with compliance, auditability, and human oversight at its core. By preserving advisor centrality, enforcing strict review processes, and delivering a clean, intuitive user experience, the platform enables scalable client engagement without compromising trust or regulatory standards.
The problem
What was slowing them down.
- 01
Limited scalability of client communication
Advisors spent significant time answering repetitive questions and preparing summaries manually.
- 02
No compliant way to use AI in client interactions
Off-the-shelf AI tools posed serious compliance risks, including unapproved investment advice and lack of audit trails.
- 03
Regulatory exposure
There was no reliable way to log conversations, track approvals, or prove compliance during audits.
- 04
Fragmented workflows
Client financial data, advisor notes, conversations, and summaries existed in disconnected systems.
- 05
Advisor visibility risks
Many AI tools risked removing the advisor from the conversation, which is unacceptable in regulated financial environments.
The solution
What we built.
Advisor-Centric AI Design
AI responses are guided by strict prompts and guardrails. The advisor's role is always visible, and human approval is required for sensitive outputs.
Three-Way Communication Model
Clients, advisors, and AI share a single conversation space with clear role labeling and shared context.
Audit-Ready Architecture
Every interaction, approval, and export is logged and timestamped for regulatory review.
Compliance-Aware AI Chat
The system automatically detects investment-specific questions and escalates them for advisor review.
Privacy-First Client Experience
Client-facing views use masked names and strict PII restrictions to maintain trust and confidentiality.
Results
What changed.
The advisory firm
Scalable client engagement
Scalable client engagement without sacrificing personalization.
Reduced advisor workload
Reduced advisor workload through AI-assisted summaries and structured workflows.
Regulatory confidence
Regulatory confidence with built-in compliance, approvals, and audit trails.
Improved client trust
Improved client trust through transparency and advisor visibility.
Market-ready SaaS foundation
Market-ready SaaS foundation suitable for long-term deployment.
AI as a competitive advantage
The firm can now confidently use AI as a competitive advantage while maintaining full compliance.
Key challenges
The hard parts, and how we solved them.
AI compliance in a regulated environment
We implemented strict prompt engineering, compliance triggers, and mandatory human review workflows.
Protecting sensitive client data
We enforced PII masking, session timeouts, encrypted data transmission, and strict data storage policies.
Preventing unauthorized investment advice
Investment-specific questions are automatically detected and escalated before any response is delivered.
Maintaining advisor centrality
UI and workflows were intentionally designed to keep the advisor visible and accessible at all times.
Usability for non-technical users
A clean, Google-inspired interface with obvious next steps ensures ease of use for clients of all ages.
Tech stack
What it runs on.
- Frontend
- Next.js/React frameworkResponsive design (mobile and desktop)Clean, Google-inspired UIReal-time chat interface
- Backend
- Node.js/Next.js API routesSecure authentication systemSession managementDatabase integration
- Database
- PostgreSQL or similar relational database
- AI Integration
- OpenAI API (ChatGPT-5.1)Custom prompt architecture with context injection
- Stores
- User accounts and authenticationClient household data (30+ fields)Complete conversation historySession metadataKnowledge base documentsAudit logs and approval trails
- Hosting & Deployment
- Vercel hosting platformGitHub repository for version control
Process
How we got there.
01Secure Authentication & Role Access
Users authenticate via email-based login with two-factor verification and a 12-hour session timeout. · Role-based access ensures that clients and advisors only see what is appropriate for their role.
02Advisor Data Entry & Context Setup
Advisors complete a structured data entry form (~30 fields) covering household profiles, balance sheets, income, expenses, action items, and narrative notes. · Advisors also manually assign Liquidity, Safety, and Growth scores. · This information is never visible to clients but is injected into AI prompts to guide responses and summaries.
03Client Dashboard & AI Interaction
Clients log in to a clean, minimal dashboard that displays their financial health at a glance. · They can chat with the AI assistant, ask questions, and request summaries while always seeing the advisor's presence.
04AI Assistance with Compliance Guardrails
The AI provides educational and planning-level information. · If a client asks an investment-specific question, the system automatically triggers advisor review and prevents unapproved responses.
05Review, Approval & Audit
Advisors review AI-generated summaries, approve client-facing exports, and maintain complete audit logs for compliance and regulatory review.
More work.
All projects
True Motion AI Chat Bot
AI-powered platform generating consistent, print-ready corrective exercise programs from assessments, enabling customization, automation, and efficient workflows for fitness professionals.
Custom Meal Planner AI Chat BOT
AI-powered chatbot delivering personalized, GLP-1-friendly meal plans with adaptive nutrition logic, automated grocery lists, and subscription-based access control system.
Building something like Plan Assist?