Finject: All-in-One CRM for Merchant Cash Advance Brokers
A specialized CRM that automates deal funding workflows for MCA brokers, replacing spreadsheets with AI-driven document parsing and lender matching.
The Challenge
MCA brokers handle complex, fast-paced deals where funding decisions must happen in hours. Traditionally, this requires manually parsing three to six months of bank statements (PDFs), calculating average daily balances and overdrafts, and manually formatting submissions to dozens of lenders, resulting in slow operations and lost deals.
- Manual document data entry takes hours and is prone to calculation errors.
- No standardized deal dashboard, causing deals to fall through the cracks.
- Lack of structured matching logic leads to submitting packages to the wrong lenders, increasing rejections.
Operations Bottleneck
Tedious manual bank statement parsing and slow, fragmented lender submissions.
Infrastructure Solved
AI-based OCR document parser and dynamic lender rule matrix.
The Solution
We built Finject—a custom SaaS CRM built specifically for the MCA brokerage workflow. The platform automates statement parsing with an AI document processor, calculates crucial risk variables, and uses a rule-based matching engine to suggest the best funding sources, enabling brokers to package and submit deals in minutes instead of hours.
- Automated bank statement parsing extracting monthly deposits, daily balances, and risk factors.
- Intelligent lender matching matrix based on historical funding rules.
- Integrated portal for merchants to securely upload files via custom single-use links.
Our Approach
Requirements Mapping
Mapped MCA broker operations to replace disparate email folders and spreadsheets with a single pipeline.
OCR Parser Design
Developed a Python FastAPI ingestion service using OCR and AI prompts to parse financial data from banking PDFs.
Deal CRM Pipeline
Built a Kanban-style pipeline in Next.js tailored for submission states (Ingested, Parsing, Lender Match, Funded).
Lender Rule Engine
Created a database of lender preferences (state restrictions, minimum monthly volume, industry blacklists).
Secure Portal
Implemented single-use client upload links using encrypted tokens to safely receive sensitive tax/bank PDFs.
Outbound Submissions
Built automated email packaging tools that bundle applications and send them directly to underwriting desks.
User Journeys
Funding Broker
Lender Underwriter
Tech Stack
Web Application
- Next.js (App Router)
- React.js
- Tailwind CSS
- Framer Motion
Backend Parser
- FastAPI (Python)
- LlamaParse / PDFPlumber
- Pandas
- PyPDF
Database & Queue
- PostgreSQL
- Supabase Auth
- Redis Key-Value Store
- Celery Workers
Messaging & Storage
- AWS Private S3
- SendGrid API
- Twilio SMS
- Firebase Cloud Messaging
Development Process
Statement Parser Ingest
Engineered parsing scripts that parse unstructured bank tables, normalizing dates, deposits, and negative balances.
Lender Matrix Engine
Designed a lightweight SQL schema for lender profile parameters allowing dynamic, compound filtering in Postgres.
Secured PDF Storage
Locked uploaded financial PDFs inside private AWS S3 buckets using short-lived signed URLs to ensure absolute compliance.
Real-time Web Sockets
Used server-sent events (SSE) to update the dashboard immediately when PDF parsing completes or email is opened.
Results & Impact
Measurable efficiency gains, reduced operations costs, and reliable integrations delivered.
Building something ambitious, or fixing something that's gone sideways?
Tell us where you are and where you're trying to get to. We'll tell you honestly whether — and how — we can help.