Case study · Firearms and ammunition supply
Hodgdon.
We built a powerful AI-driven financial and inventory management chatbot that allowed a firearms and ammunition supplier to monitor, analyze, and predict business performance in real-time.
- What we built
- AI chatbot for financial and inventory management
- Stack
- Next.js, Tailwind CSS, Next.js (with Vercel AI SDK)
- Services
- AI chatbotPredictive analyticsData integration
About the project
What is Hodgdon?
We built a powerful AI-driven financial and inventory management chatbot that allowed a firearms and ammunition supplier to monitor, analyze, and predict business performance in real-time. The solution integrated financial data, sales metrics, and inventory management insights into a conversational AI agent, enabling decision-makers to simply “ask” the chatbot about revenue, costs, inventory levels, shipments, or predictive forecasts, without needing to dig into spreadsheets or dashboards.
The problem
What was slowing them down.
- 01
Large master sheets
The team relied on large master sheets with thousands of rows, updated through MS Dynamics, making it hard to track live changes.
- 02
Scattered data
Sales, inventory, and finance data were scattered across multiple sources.
- 03
Heavy forecasting analysis
Forecasting (predicting inventory demand, revenue, and cashflow trends) required heavy analysis that most team members couldn’t perform easily.
- 04
Manual reporting
Manual reporting consumed significant time and delayed decisions.
The solution
What we built.
Conversational AI agent
Users could ask questions like “What’s the current inventory value?” or “Show me this month’s revenue vs. shipment costs” and get instant answers.
Predictive analytics
AI generated forecasts on revenue, inventory needs, and order dispatch timelines.
Role-based access
Custom user permissions ensured data security while allowing team-wide usage.
Real-time data sync
Integrated directly with MS Dynamics to fetch and update live financial and operational data.
Comprehensive reporting
The chatbot provided insights into cash flow, account receivables, payables, shipment costs, order values, and stock levels.
Results
What changed.
Finance, sales, and operations teams
75% faster decision-making
75% faster decision-making compared to manual reporting.
Real-time financial transparency
Real-time financial transparency, reducing errors and delays.
Improved operational efficiency
Improved operational efficiency by automating daily queries about stock, costs, and shipments.
Forecast-driven planning
Forecast-driven planning reduced overstocking and optimized cash flow.
High adoption rate
High adoption rate across finance, sales, and operations teams since the solution was as simple as asking questions in plain English.
Key challenges
The hard parts, and how we solved them.
Handling complex financial queries
Implemented advanced AI prompt engineering and code interpretation via E2B to deliver precise answers.
Real-time data synchronization
Built pipelines with MS Dynamics and Tinybird to ensure live updates.
Ensuring data security
Role-based permissions and encrypted data access were integrated at all levels.
User adoption
Designed chatbot UX with a simple, natural conversation flow to make adoption frictionless.
Tech stack
What it runs on.
- Frontend
- Next.jsTailwind CSS
- Backend
- Next.js (with Vercel AI SDK)
- AI
- OpenAI (with E2B for code interpretation)
- Observability & Tracing
- Langfuse
- Database
- MongoDB (user details)Tinybird (core/large-scale data)MS Dynamics (data fetching and syncing)
- Session Management & Analytics
- PostHog
Process
How we got there.
01Discovery & Analysis
Identified pain points in financial reporting and inventory management · Mapped required data points from MS Dynamics
02Data Integration
Connected MS Dynamics, MongoDB, and Tinybird pipelines · Synced financial, sales, and inventory metrics in real-time
03AI Agent Development
Built chatbot interface with Next.js + Tailwind CSS frontend · Used Vercel AI SDK + OpenAI for conversational intelligence and code interpretation
04Testing & Optimization
Implemented observability via Langfuse · Used PostHog for session management and analytics
05Deployment & Training
Rolled out AI chatbot across departments · Conducted user onboarding and workflow training
More work.
All projects
Rufus Labs
AI-powered warehouse system that analyzes wearable scanner data to track worker productivity, optimize item placement, and deliver real-time insights.
Gathrly
AI-powered photo booth SaaS that generates instant stylized photos and videos for events, with automated delivery, payments, and venue revenue sharing.
Building something like Hodgdon?