An autonomous AI agent that investigates incidents while you sleep. Self-hosted. APRA CPS 234 ready. BYO LLM keys. Your data never leaves your network.
Financial services face unique constraints when investigating production incidents.
Sending production logs, database queries, or code to cloud AI services violates APRA guidelines and PCI DSS requirements. Most incident investigation tools require cloud connectivity.
Payment processing incidents cost $5,600/minute on average. Waiting hours for senior engineers to manually correlate logs, code, and database state is unacceptable.
SupportAgent solves both problems: AI-powered investigation speed with 100% on-premises data sovereignty.
Every feature designed with APRA, PCI DSS, and SOC 2 requirements in mind.
100% on-premises deployment. Customer PII, transaction data, and audit logs stay within your infrastructure. Zero external API calls for sensitive data.
Bring your own API keys for OpenAI, Anthropic Claude, Google Gemini, or run fully offline with Ollama. Your keys, your control, no vendor lock-in.
Meets information security requirements for Australian ADIs. Full audit trails, access controls, and data classification support.
Self-hosted architecture means cardholder data environments remain isolated. No third-party data transmission for investigations.
Real scenarios where SupportAgent accelerates incident resolution.
Investigate failed transactions, gateway timeouts, and reconciliation errors across your payment stack.
Example: AI traces a failed payment from API gateway through payment processor integration to identify a misconfigured timeout setting.
Correlate market data feeds, order management systems, and execution engines to find root causes fast.
Example: When orders fail during market open, AI searches code, database queries, and recent deployments to identify the bottleneck.
Investigate account balance discrepancies, interest calculation errors, and batch processing failures.
Example: AI analyzes overnight batch jobs, identifies the SQL query causing incorrect balance updates, and provides the exact fix.
Trace why reports to ASIC, APRA, or ATO failed and identify data quality issues at source.
Example: AI correlates data lineage from source systems through transformation to identify where regulatory report data was corrupted.
Investigate across all systems from one platform.
Same pricing for all industries. No FinTech premium.
Enterprise/Private Cloud options available. Contact us
SupportAgent is designed for APRA CPS 234 compliance. As a self-hosted solution, all data remains within your network perimeter. It provides full audit logging, role-based access control, and supports your information security classification requirements. Your compliance team can review the complete architecture documentation.
SupportAgent connects to databases in READ-ONLY mode only. It cannot modify, insert, or delete data. All queries are logged for audit purposes. You control exactly which databases and tables the AI can access through granular permissions.
No. When using the local LLM option (Ollama), zero data leaves your network. Even the AI model runs locally. If you choose to use OpenAI or other cloud LLMs, you can configure exactly what data is sent - typically just sanitised error messages, not raw customer data.
SupportAgent includes configurable data masking for PII, card numbers, and other sensitive data. You define patterns to mask before any AI analysis. Audit logs track what data was accessed but store masked versions only.
Yes. We provide complete architecture documentation, data flow diagrams, security controls specification, and network topology for your security team's review. This documentation supports APRA, PCI DSS, and SOC 2 compliance assessments.
15-day free trial. No credit card. Runs entirely on your infrastructure.