Security & Reliability

Built to be trusted with production

Bring Your Own Key (BYOK)

Bring your own LLM key, with multi-vendor multi-model support.

Data Leakage Prevention

PII and Sensitive information masking with strict data governance layers.

Multiple Deployment Options

On-prem, Air-gapped and Private Cloud, Managed Cloud SaaS Deployments, configured and validated by a dedicated Forward Deployment Engineer.

Access Control and Auditing

RBAC, MFA, Custom SSO/SCIM provisioning and Audit trails

Compliant on

SOC 2GDPRDPDPAEU AI ActHIPAA

Deploys to

Self-hosted VPCOn-premPrivate cloudAir-gappedCloud SaaS

Security applied in layers

TLS termination happens at your ingress controller or reverse proxy. The application itself sends Strict-Transport-Security with a 2-year max-age, including on subdomains, so browsers force HTTPS on every future visit.

Uptime & Reliability

Open source's benchmarks, reproducible today

0ms

Webhook ingestion p99

threshold < 200ms

0

Peak RPS tested

k6 burst test, ramp to 500 VUs

0

5xx errors at peak

clean 429s under flood

0s

DB failover RTO

Patroni chaos test

Automatic Postgres failover, step by step

patroni-1 becomes unreachable to the rest of the cluster.

Benchmark results

Every number below is reproducible with the load-test and chaos scripts in the repo.

MetricResultTarget
Webhook sustained p99 (50 VUs, 5 min)< 10 ms< 200 ms
Webhook burst peak RPS (ramp to 500 VUs)3,917 RPS
5xx errors during burst00
DB failover RTO (Patroni chaos test)11 s< 60 s
Redis failover RTO (chaos test)5 s< 10 s
In-flight request loss on deploy0 dropped0
Accuracy benchmarks

80%

Accurate

Integrated with the production tools.

Human in the loop

Every RCA and runbook reasoning undergoes Quality check with manual approval before it enters the library.

Steering capability

Edit and redirect the reasoning and tool-usage steps, then re-execute to sharpen the result.

Guardrails

Hard limits on reasoning loops, tool-use depth and iterations, token usage and LLM response bounds.

Redaction

PII and Sensitive information masking. Strict data governance preventing data leakage to LLM.

Evaluation benchmarks

Accuracy gates with confidence scores and validation benchmarks.

Shadow mode

Simulate against the live system with no write provision. Runbooks and playbooks move to live only once approved.

From incident timeline to post-mortem draft

The incident timeline closes - every ack, escalation, comment, and status change is already recorded in order.