Workflow directory
Everything we've built and everything we're building next. See one you need? Schedule it — we'll build it and have it live within 24 hours.
Coming soon
In the pipeline. Schedule one and we'll build it in the next 24 hours — you get an email the moment it's live.
MSP Quote-to-Cash AR Anomaly & Dunning Engine with Human-in-the-Loop Governance (Kaseya + Stripe + QuickBooks + Gmail)
The missing server fetches open quotes from ConnectWise CPQ and unpaid invoices from /stripe, then uses reasoning to detect quote-to-invoice mismatches, flag overdue accounts by aging bucket (30/60/90 days), segment customers by risk tier, and draft personalised dunning emails. The output is a prioritised collections queue with anomaly flags delivered to the MSP's ops team, closing the full quote-to-cash loop without any manual AR reconciliation.
- Google OAuth2
- Microsoft 365
- Oakallow dashboard
- MFA systems
- SQLite
- OAuth 2.1
- +2
Persistent UK Property Risk Knowledge Graph Agent with Government Open Data, Land Registry Price, Walkability, Satellite Land Hazard & Live Web Intelligence Enrichment
The missing server stress-tests a UK mortgage portfolio and produces structured property-level risk classifications and narrative reasoning. The Memory workflows stores each property's risk history, area intelligence scores, and breach narratives as a persistent knowledge graph, allowing the agent to reason across time ('this property has been on watch for 3 consecutive months') and across the portfolio ('all breaches cluster in postcodes with declining area scores'). On each monthly run, the agent enriches the graph with new stress results and surfaces cross-portfolio patterns — such as correlated geographic risk concentrations — that a single-pass analysis would miss.
- HM Land Registry Price Paid Data
- Pipeworx gateway
- Xvfb
- SQLite
- Docker/Podman
- Walk Score API
- +2
Stripe-to-QuickBooks ASC 606 Multi-Currency Revenue Close Automation
The missing server pulls nightly Stripe invoice, subscription, and refund events via /stripe, applies ASC 606 / IFRS 15 reasoning to classify each transaction and generate double-entry journal entry drafts, then persists a full audit trail and variance log to a local SQLite database via /sqlite for human review before final ledger push. The SQLite layer acts as a staging and reconciliation store, enabling period-over-period variance queries, rollback, and audit export without requiring a live accounting system connection at every step.
- SQLite
- QuickBooks
- ECB (European Central Bank)
- Stripe
- Solana blockchain (MINT Protocol)
- QuickBooks Online API
Bitcoin AML Risk Memory, SAR Autopilot & Entity/Financial Identifier Verification
The Maestro workflows streams live Bitcoin transaction data for monitored wallets; the compliance applies FATF/AML heuristics and reasoning to score wallet risk and draft FinCEN SAR narratives. The Memory persists a knowledge graph of wallet entities, prior risk scores, counterparty clusters, and case history so that each new transaction is evaluated in the context of the full investigative record — enabling pattern detection across time that a single-pass analysis would miss.
- EU VIES
- DNS/MX
- EU EOS Customs
- US state business registries
- Bitcoin
- Stripe
- +2
Algo Risk Audit Pipeline with Time-Series Storage
The missing server fetches backtest results from /quantconnect and FRED macro indicators, synthesises a plain-English risk narrative, and writes structured risk memos plus breach flags into a ClickHouse time-series database via /clickhouse. Over time, ClickHouse accumulates a longitudinal audit trail of every strategy's risk profile across macro regimes, enabling compliance teams to query historical risk posture and demonstrate regulatory due diligence on demand.
- QuantConnect API
- QuantConnect
Data cleaning and reconciliation before close
Before month-end close, accounting teams spend hours cleaning messy data, reconciling discrepancies between systems, and auditing entries. Data quality issues are discovered late, delaying close cycles and requiring rework.
Stripe + Supabase Workflow
Supabase databases integrate with Stripe's payment infrastructure to automatically recover failed payments and optimize transaction processing. Payment events from Stripe are logged and tracked in Supabase, enabling dispute recovery and Link payment optimization.
Manual receipt and invoice data extraction
Finance teams manually pull data from bank statements, receipts, and invoices into spreadsheets. This involves reading unstructured documents and manually entering or categorizing transaction details, consuming hours of effort per month.
Transaction categorization and account matching
Bookkeepers spend 80% of their time on data entry and manual transaction categorization, deciding which account each transaction belongs to and whether it matches existing records. This repetitive judgment work delays analysis and introduces errors.