Platform
Live Runs
Every flow, every run, fully observable. Live status, node-level execution detail, automatic retries, and an immutable record of what happened.
RUN HISTORY
See every run that touched production.
Every run gets a unique Run ID and a complete record. One screen, every flow, what happened today.
- Confidence + business-rule routing. Field-level thresholds combine with rules like missing docs, risk appetite, and duplicates.
- Real-time status. Status indicators update as runs progress, finish, fail, or pause for human review.
- Filter and search. By flow, environment, status, trigger source, date range, or Run ID.
- Trigger attribution. Manual upload, HTTP API call, or Outlook email pull. Every run knows where it came from.
Node-level visibility
Drill into any run, see every step.
Open a run and the Canvas lights up with the actual path that executed. Hover any node for inputs, outputs, model invoked, duration. Debugging stops being guesswork.
- Per-node telemetry. Which InsurGPT™ model was invoked, retries, the exact branch decision, the result.
- Selective recompute. Fix one bad extraction and rerun just that node. No need to start the flow over.
- Inspect variables and state. Every variable, every secret reference, every workflow input. Secret values stay masked.
Built For Production
Resilient by default. Recoverable by design.
Real-world failure modes are expected and handled. When something does go wrong, recovery is a click, not a ticket.
- Automatic retries on transient failure. Network blips, downstream timeouts, and rate limits resolve without operator involvement.
- Idempotent execution. Same input, same Run ID, same result. Safe to retry, safe to integrate with at-least-once delivery systems.
- State engine survives restarts. A flow paused at HITL last week resumes cleanly when the reviewer submits today.
FEATURE INDEX
Inside Live Runs.
Run history & observability | Capability
Every flow execution gets a unique Run ID and a full record. One screen shows what ran today, where it came from, and how it performed.
- Run ID for every execution
- Real-time status indicators on every run
- Trigger attribution: manual upload, HTTP, Outlook email
- Filter by flow, environment, status, date, Run ID
- Duration, throughput, and SLA tracking
- Exportable run history for reporting
Node-level execution detail | Capability
Open any run and the Canvas lights up with the actual execution path. Hover any node for inputs, outputs, model invoked, and duration. Debugging stops being guesswork.
- Visual run replay on the Canvas
- Per-node inputs, outputs, duration
- Model invocation traces for every InsurGPT™ call
- Branch decision capture for If/Switch nodes
- Loop iteration breakdown
- Variable inspector with masked secrets
- HITL pause and resume timestamps
Reliability & recovery | Capability
Real-world failure modes are expected and handled. Transient errors retry automatically, idempotent execution prevents duplicate work, and the state engine survives restarts.
- Automatic retries on transient failure
- Idempotent execution with Run ID deduplication
- Production-grade state engine
- Selective recompute at the node level
Governance & audit | Capability
Every run is logged immutably with actor, action, resource, and timestamp. Flow versions tie back to every execution, secrets stay masked, and exports satisfy auditors.
- Immutable run log for every execution
- Actor, action, resource, timestamp on every event
- Secret values masked in run logs
- Flow version tied to every run for rollback safety
- Environment isolation between draft, staging, production
- Compliance export for SOC 2, ISO 27001, HIPAA, NYCRR 500
- RBAC enforced on every run view and action
Resources & insights
More on Human-in-the-Loop
Research
Page stream segmentation with LLMs
How Bevaya Labs approaches a foundational problem in insurance document AI.
Case Study
Workers' comp carrier processes claims 100x faster
How indexing automation delivered 432% ROI in 12 months.
Architecture
Inside the Bevaya platform architecture
How specialized models, HITL controls, and integrations come together in production.
More Capabilities
Explore the rest of the platform.
Designed, deployed, and governed together. Powered by InsurGPT™ and accessed through the AI Assistant.
Workflow Canvas
Visual builder and production runtime for every automation.
Current page ReviewHuman-in-the-Loop
Configurable review queues with X-Ray verification and a patented feedback loop.
Current page DocumentsDocument Intelligence
Read any insurance document — hundreds of carrier formats, scanned or digital.
Current page GroundingGrounded Explainability
Every value traceable to its source. X-Ray Highlight Mode brings citations to reviewers.
Current page AnalyticsAnalytics Dashboard
Live accuracy, STP rates, reviewer SLA, and agent performance across every workflow.
Current page GovernanceGoverned Automation
Immutable audit trails, role-based access, flow versioning. Compliance is the architecture.
Current pageTrust & Security
Trust by design
Built for an industry where data security isn't optional.
Your data stays yours.
Never shared with other customers or vendors. Bevaya doesn't train shared models on your data.
Visit the Trust CenterYour data · only your team sees it
Encrypted end-to-end.
256-bit AES encryption, in transit and at rest. Independent third-party audits conducted annually.
Visit the Trust CenterAudited annually · independent third party
Runs in Azure.
Enterprise-grade infrastructure, hosted where insurance organizations already trust their data.
Visit the Trust CenterDeploy where your stack already lives
Every decision audited.
Immutable audit logs. Confidence scoring. Human-in-the-Loop review on low-confidence items.
Visit the Trust CenterImmutable trail · every decision, every reviewer
FAQ
Common questions.
Run ID, flow name and version, environment, trigger source, actor, start and end timestamps, duration, branch path taken, every node executed with inputs and outputs, every model invocation, every retry, and final status. Logs are immutable and exportable.
Transient failures retry automatically. The Run ID stays the same across retries, so observability and audit are preserved. Persistent failures surface to operators with the full failure context attached.
When an external system calls a Bevaya HTTP trigger with the same input twice, Bevaya treats it as the same run and does not create duplicate work items. The Outlook email trigger applies the same protection, so a re-delivered email does not produce a duplicate flow execution.
No. Secret values are masked in every UI view and redacted from every run log. The Run inspector shows the reference (e.g., "GUIDEWIRE_PROD_TOKEN") but never the value.
Bevaya is purpose-built for insurance AI workflows. Runs surface InsurGPT™ model invocations, HITL pause and resume events, and field-level confidence in the same view as durations and branch paths. A generic workflow engine logs steps. Bevaya logs the insurance decisions inside each step.
Get Started
Ready to design, deploy, and govern your AI Workforce
Bevaya AI Agents can help you triage, analyze, and recommend across underwriting, claims, and policy servicing. Let's connect and show you how it works.