Service Level Agreement

1. Scope of the Agreement

This Service Level Agreement ("SLA") outlines the performance metrics, uptime guarantees, and latency thresholds provided by HarvestGroup360 to its institutional partners. This agreement is purely a technical specification and does not involve monetary deposits, brokerage clearing, or financial compensation.

2. 99.99% Uptime Guarantee

HarvestGroup360 commits to a Monthly Uptime Percentage of at least 99.99% for its core infrastructure, including the REST API, WebSocket streams, and the order matching engine.

Definition of Downtime: Downtime is defined as a period of at least one (1) consecutive minute during which the API returns a 5XX server error to all valid requests, or the WebSocket endpoint completely refuses connections. Temporary latency spikes or client-side network routing issues do not constitute downtime.

Exclusions: Downtime does not include (a) Scheduled Maintenance, provided 48 hours notice is given, (b) Force Majeure events, or (c) suspensions resulting from algorithmic abuse or violation of API rate limits.

3. Latency and Execution Metrics

In high-frequency quantitative environments, predictability is more important than raw speed. We provide the following internal processing guarantees:

• Order Acknowledgment: For clients utilizing physical cross-connects in NY4/LD4, HarvestGroup360 guarantees an internal processing latency (time from packet ingress at our switch to algorithmic acknowledgment) of under 150 microseconds for 99% of requests.
• Market Data Jitter: We utilize hardware timestamping at the network interface card (NIC) level. Timestamp jitter on tick data is guaranteed to be under 5 microseconds.

4. Incident Response and Resolution

Our engineering team operates 24/7/365 to ensure infrastructure stability. Incidents are categorized by severity:

• Severity 1 (Critical): Complete failure of API or WebSocket infrastructure. Response Time: < 15 minutes.
• Severity 2 (High): Significant latency degradation or partial data feed interruption. Response Time: < 30 minutes.
• Severity 3 (Standard): Historical data query anomalies or general technical inquiries. Response Time: < 4 hours.

5. Data Integrity and Redundancy

HarvestGroup360 utilizes synchronous replication across three separate geographic availability zones. In the event of a catastrophic failure at our primary data center, algorithmic traffic is automatically failed over to a secondary cluster within 3 seconds. No historical tick data will be lost during this transition.