Back to projects

Full Stack · GenAI · Engineering case study

Fund Flick Lending Operations Suite

Client: Fund Flick · Maitrii Infotech

A lending and HR SaaS platform with LangGraph AI agents, a web dashboard, and an Android field app for loans, collections, payroll, attendance, and leave.

React NativeNext.jsNode.jsFastAPIMongoDBLangGraphAWS Lambda
Illustrative architecture
ClientsAPIsModulesStorage & servicesData · MongoDBAWSOBSERVABILITYattendancerecordslogsAPMtracesNext.jsWeb dashboardFund Flick appReact Native · AndroidNode.js · FastAPIServices onAWS LambdaLendingOrigination · collectionstasksHR & payrollPeople · Payrolloutward paymentsTime & leaveAttendance · leaveAmazon S3Letters · uploads · exportsPush notificationsTasks · follow-upsAI agentsLangGraph · LangChainMongoDBLoans · collectionsemployees · payrollCloudWatchLogs · metricsDatadogAPM · dashboardsLangSmithAgent traces

01

The problem & context

Fund Flick, by Maitrii Infotech, is a lending operations suite for LAP, MSME, vehicle, business, personal, and gold lenders. One SaaS platform covers lending work and the HR suite, with AI agents for HR, on the web and in the Fund Flick app for field officers, collection agents, and managers.

02

Architecture

Illustrative view of the Fund Flick platform. A Next.js web dashboard and the React Native Fund Flick app call Node.js and FastAPI services on AWS Lambda. Lending modules handle loan origination, collections, and tasks, with push notifications to field staff, while the HR suite manages people, payroll with outward payments, attendance, and leave. Records live in MongoDB, HR letters, uploads, and exports in Amazon S3, and LangGraph agents add HR assistance traced in LangSmith. CloudWatch and Datadog monitor the platform.

Illustrative system view, not a production screenshot.

Illustrative architecture
ClientsAPI · FastAPIAgents · LangGraphModels & storageData · MongoDB AtlasAWSOBSERVABILITYrecordsvector searchreportstraceslogs · APMNext.jsHR dashboardFund Flick appAndroid · staffFastAPI λHR APIsAWS LambdaLangGraph agentsPolicy Q&A · lettersinsights · parsingRetrieverAtlas Vector SearchLLMvia LangChainAmazon S3Letters · uploadsMongoDB AtlasEmployees · payrollattendance · leaveVector indexHR policy chunksLangSmithAgent traces · evalsCloudWatch · DatadogLogs · metrics · APM

02.1

Architecture: HR AI agents

Illustrative view of the Fund Flick HRMS. A Next.js dashboard and the Fund Flick Android app call a FastAPI backend running on AWS Lambda, which stores employees, payroll, attendance, and leave in MongoDB Atlas. LangGraph agents built with LangChain answer HR policy questions by retrieving policy chunks through MongoDB Atlas Vector Search, draft HR letters stored in Amazon S3, summarize payroll and attendance reports, and parse uploaded resumes and onboarding documents. LangSmith traces the agents, and CloudWatch and Datadog monitor the platform.

Illustrative system view, not a production screenshot.

03

Implementation & my contribution

  • Built the React Native field app for collection cases, EMI repayment records, field visits, and tasks, with selfie and GPS proof of each visit.
  • Added push notifications for task assignments, comments, and collection follow-ups.
  • Built the Next.js dashboard across lending and the HR suite: Employees, HR Letters, Payroll, Outward Payments, Reports, Attendance, Leave Management, Deletion Requests, and Settings.
  • Developed Node.js and FastAPI services on AWS Lambda, with MongoDB Atlas for loans, collections, employees, payroll, attendance, and leave.
  • Built LangGraph agents with LangChain: an HR policy assistant over MongoDB Atlas Vector Search, HR letter drafting, payroll and attendance insights, and resume and onboarding document parsing.
  • Stored HR letters, uploads, and report exports in Amazon S3, traced agents with LangSmith, and monitored the platform with CloudWatch and Datadog.
  • Built the fundflick.in marketing website in Next.js with GSAP animations.
04

Outcome

Delivered a live lending and HR SaaS platform with AI-assisted HR workflows on the web and Google Play, plus the public marketing website.

Testimonials

Voices from collaboration.