JJA.CO
Senior AI Engineer – Document / Invoice Processing with ERP Integration
Los Angeles | San Francisco
Building the reserve layer of machine finance. We unlock trapped working capital for small and mid‑sized businesses by transforming invoices into programmable, tradeable financial assets. We bypass legacy bottlenecks in banking and ERP systems, providing institutional investors with direct access to short‑term receivables as a new source of yield. Our infrastructure is live, processing real corporate invoices today, and designed to scale into a global liquidity rail connecting enterprises, AI agents, and capital markets.
The Role Build production AI agents that extract structured financial data from enterprise systems at scale. You’ll integrate with SAP, Oracle, and NetSuite; process 10K+ invoices monthly with 95%+ accuracy; and orchestrate multi‑agent workflows that convert unstructured documents into blockchain‑ready assets.
Must Haves
LLM Development
Document or Invoice Parsing/Processing or Data Extraction from Documents
ERP Integration
Technical Requirements
Production Multi‑Agent Systems
Built systems processing 1M+ monthly requests using LangChain or LangGraph
Achieved 50%+ cost reduction through token optimization, caching, or multi‑model routing
Implemented parallel agent execution with state persistence and error recovery
95%+ field extraction accuracy on PDF invoices, scanned documents, and complex layouts
Production experience handling OCR artifacts, malformed documents, and multi‑format parsing
Real‑time validation pipelines with GST/VAT checks and automated reconciliation
ERP Integration Experience
Native connectors for SAP (S/4HANA, ECC), Oracle (NetSuite, ERP Cloud), or Microsoft Dynamics
Real‑time data sync between LLM systems and business systems (sub‑second latency)
Complex data transformation between unstructured documents and ERP schemas
Agent Runtime Infrastructure
E2B or Daytona implementation for secure agent code execution
Container orchestration achieving 20x+ throughput improvements
SOC2/GDPR compliance in financial data processing systems
Must‑Have Experience
Accuracy proof: 95%+ extraction rates on production financial documents
Integration proof: 5+ external APIs with authentication, webhooks, and rate limiting
Compliance proof: Experience with SOX, audit trails, or other financial regulatory frameworks
Technical Stack Processing: Pydantic, structured outputs, function calling, semantic parsing
Ideal Background
3+ years building production LLM applications with measurable performance metrics
Open‑source contributions to LangChain or Composio (1000+ commits preferred)
Enterprise software integration at fintech or AI infrastructure companies
Seniority Level Mid‑Senior level
Employment Type Full‑time
Industries Software Development
#J-18808-Ljbffr
Building the reserve layer of machine finance. We unlock trapped working capital for small and mid‑sized businesses by transforming invoices into programmable, tradeable financial assets. We bypass legacy bottlenecks in banking and ERP systems, providing institutional investors with direct access to short‑term receivables as a new source of yield. Our infrastructure is live, processing real corporate invoices today, and designed to scale into a global liquidity rail connecting enterprises, AI agents, and capital markets.
The Role Build production AI agents that extract structured financial data from enterprise systems at scale. You’ll integrate with SAP, Oracle, and NetSuite; process 10K+ invoices monthly with 95%+ accuracy; and orchestrate multi‑agent workflows that convert unstructured documents into blockchain‑ready assets.
Must Haves
LLM Development
Document or Invoice Parsing/Processing or Data Extraction from Documents
ERP Integration
Technical Requirements
Production Multi‑Agent Systems
Built systems processing 1M+ monthly requests using LangChain or LangGraph
Achieved 50%+ cost reduction through token optimization, caching, or multi‑model routing
Implemented parallel agent execution with state persistence and error recovery
95%+ field extraction accuracy on PDF invoices, scanned documents, and complex layouts
Production experience handling OCR artifacts, malformed documents, and multi‑format parsing
Real‑time validation pipelines with GST/VAT checks and automated reconciliation
ERP Integration Experience
Native connectors for SAP (S/4HANA, ECC), Oracle (NetSuite, ERP Cloud), or Microsoft Dynamics
Real‑time data sync between LLM systems and business systems (sub‑second latency)
Complex data transformation between unstructured documents and ERP schemas
Agent Runtime Infrastructure
E2B or Daytona implementation for secure agent code execution
Container orchestration achieving 20x+ throughput improvements
SOC2/GDPR compliance in financial data processing systems
Must‑Have Experience
Accuracy proof: 95%+ extraction rates on production financial documents
Integration proof: 5+ external APIs with authentication, webhooks, and rate limiting
Compliance proof: Experience with SOX, audit trails, or other financial regulatory frameworks
Technical Stack Processing: Pydantic, structured outputs, function calling, semantic parsing
Ideal Background
3+ years building production LLM applications with measurable performance metrics
Open‑source contributions to LangChain or Composio (1000+ commits preferred)
Enterprise software integration at fintech or AI infrastructure companies
Seniority Level Mid‑Senior level
Employment Type Full‑time
Industries Software Development
#J-18808-Ljbffr