Show HN: Delfyn, AI Agent to help businesses get paid faster
I'm Junior, founder of Delfyn, and I’ve been building an AI-powered agent that helps finance teams get paid on time. If you've ever dealt with slow-paying B2B customers, broken invoice matching, or cash flow uncertainty, this might resonate.
The problem: In B2B, waiting 60–90 days to get paid is common. AR teams spend hours manually:
- Reconciling payments to invoices
- Chasing customers
- Updating spreadsheets
Explaining why cash flow projections keep being wrong
Meanwhile, late payments quietly bankrupt 1 in 4 SMEs in Europe.
What Delfyn does: We built an AI agent that helps automate:
Invoice-to-payment matching using rules, embeddings (e.g. “0099” → “INV-2023-0099”), and retrieval models
Cash flow forecasting with transaction-level granularity
Proactive reminders + incentive recommendations to reduce DSO
We're combining deterministic logic (amount, IBAN, date) with semantic models (e.g. e5-base-v2) and LLMs (for gray-zone explanations only).
Who it's for: Finance teams (esp. in B2B SaaS, agencies, professional services)
Founders managing AR themselves
Companies tired of relying on NetSuite exports + wishful thinking. Or even worst sending payments collections emails to customers who have already paid. I've seen that a lot.
No comments yet