Product

Explore Veras

Follow one operational situation from first understanding to organizational learning, and see how Veras supports each step with evidence, explicit uncertainty and clear operational authority.

Veras complements the systems operators already use. It brings together relevant signals, work history, procedures and prior experience to support operational reasoning.

Veras is in active development. This walkthrough illustrates our evolving product vision.
How understanding is established

Operational understanding built on evidence, knowledge and experience

  • Operational Reality What is happening now: assets, operating conditions, measurements and operator observations.
  • Technical Knowledge What should happen: manuals, constraints and procedures.
  • Historical Experience What happened before: incidents, work history and operator experience.
Operational context

Reasoning

Identifies the context that matters, connects relevant evidence and knowledge, and develops an assessment that makes its reasoning and uncertainties explicit.

Veras starts with the information already available and becomes richer as it is used.

Veras makes supporting evidence, conflicting evidence and unresolved questions explicit, while keeping operational decisions with the designated authority.

01 Situation

What is happening?

Brings together the signals, dependencies, recent activity and knowledge relevant to the situation.

  • Connects the current condition to assets and operational dependencies
  • Brings relevant signals, recent activity and applicable knowledge together in one place
  • Shows current risk and operational impact together
Situation briefing Critical Cooling Loss, Data Hall A
Degradation in progress

What is happening?

Situation now

Cooling capacity is degraded, thermal margin continues to narrow, and service performance is beginning to be affected.

Thermal margin Narrowing
Thermal window remaining ~45 min to defined
thermal limit
Service performance Degrading
What we know
  • Degradation has persisted for ~20 min
  • CU‑04 is delivering less cooling than expected
  • Data Hall A thermal conditions continue to deteriorate
  • Dallas recovery site is available
What remains uncertain
  • Failure mechanism within CU‑04
  • Time required for local recovery
  • Viability of local service redundancy under the same thermal exposure
Relevant dependencies
Electrical supply Power availability
→
Control & actuation Commands, sensing, actuation
→
CU‑04 Cooling delivery
→
Data Hall A thermal condition Thermal environment
→
Critical equipment Thermal exposure
Power & control Supply, commands and sensing
→
CU‑04 Cooling delivery
→
Data Hall A Thermal condition and critical equipment
Available resilience
Local service redundancy Same exposure
Dallas recovery site Capacity available
Recent relevant activity
Yesterday at 18:40 CU‑04 maintenance: air filter replaced; compressor service and inspection completed

Relevant context, not yet linked to cause.

View maintenance record
Relevant knowledge Potentially applicable

3 relevant procedures and guidance sources identified for this situation.

Includes cooling-loss response, cross-site failover and OEM guidance.

View relevant knowledge
Investigation starting point

Cooling degradation is confirmed; the cause is not. First determine whether the issue is isolated to CU‑04 or extends beyond it.

02 Investigation

What uncertainty should we reduce?

Test competing explanations without forcing an early conclusion.

  • Keep competing explanations visible and clear while the cause remains unresolved
  • Organize evidence, uncertainty and limitations so the investigation does not have to be reconstructed manually
  • Prioritize the information that most changes the next decision
Investigation Critical Cooling Loss, Data Hall A · Investigation in progress
Hypotheses under assessment: 3
Symptom Degradation for ~20 min • CU‑04 was serviced yesterday • Local service redundancy shares the same thermal exposure

What is driving the cooling degradation?

Hypotheses under assessment

Each explanation shows why it is plausible and what limits it. Nothing is concluded early.

Hypothesis A

Mechanical or refrigerant‑circuit degradation within CU‑04

Why plausible Cooling delivery is degraded and compressor service occurred recently.
Limitation Recent maintenance does not establish the current failure mechanism.
Hypothesis B

Internal CU‑04 control, sensing or actuation fault

Why plausible The unit may remain available while not operating at the commanded state.
Limitation Current evidence has not established an internal control or instrumentation fault.
Hypothesis C

External upstream condition affecting CU‑04 operation

Why plausible Power supply or external commands could reduce CU‑04 performance without an internal equipment failure.
Limitation No abnormal upstream condition has yet been established.
Information needed next

Confirm CU‑04 cooling output and whether the compressor is operating as expected

Why this matters Distinguishes internal cooling‑delivery degradation from a control or upstream condition and changes which recovery paths remain viable.
Relevant evidence, guidance and history
Cooling Loss, Data Hall B, 14 months ago Similar previous incident
Compressor control failure; response: local failover and cooling isolation. View investigation report
CU‑04 maintenance record Yesterday at 18:40
Air filter replaced; compressor service and inspection completed. View maintenance record
CU‑04 Thermal Protection & Recovery OEM guidance, Rev. 3
CU‑04 guidance triggers protective response above 27°C representative inlet temperature if rising. View OEM guidance
Cooling Loss Emergency Response Approved procedure, Rev. 6
Escalation required if expected local recovery exceeds the remaining thermal window. View emergency response procedure
03 Decision & Execution

What should we do, and is it authorized?

Turn the latest evidence and operating constraints into a decision that can be reviewed and authorized.

  • Compare recovery paths against the latest operating state
  • Surface the applicable procedure, prerequisites and required authority at the point of action
  • Keep execution requirements and evidence connected to the chosen path
Decision & Execution Critical Cooling Loss, Data Hall A · Recovery path
Authorization pending

What should we do, and is it authorized?

Investigation synthesis
Assessment updated after new evidence

Checks confirmed degraded CU‑04 cooling output and found no shared upstream abnormality. Local restoration is not expected within the remaining thermal window; root cause remains unresolved.

Compare recovery paths
A Recover locally
Strength Avoids service transfer
Constraint Restoration timing remains uncertain
B Local service failover
Strength Preserves service locally and avoids cross‑site transfer
Constraint Local service redundancy shares the same thermal exposure

Because both paths remain exposed to the same cooling condition, local failover may preserve service temporarily but does not materially reduce the thermal risk.

C Remote failover to Dallas
Suggested for review
Strength Removes critical workloads from the current thermal exposure
Constraint Requires transfer and replication readiness

Local restoration timing remains uncertain while the thermal window continues to narrow. Dallas removes critical workloads from the current thermal exposure once transfer readiness is confirmed.

Authorization gate
Authorization pending

Veras has suggested a path. The workflow remains pending until the designated authority approves execution.

Required authority Designated incident authority
Before execution
Confirm transfer readiness
Confirm applicable failover procedure and prerequisites
Complete required operational notifications
Record authorization
Execution package
Cross‑Site Service Failover Approved procedure, Rev. 8
Applicable procedure, prerequisites, execution owner and evidence required to verify the outcome. View procedure
04 Verification

Did it actually work?

Verify the operational state, not just that an action was completed.

  • Compare the observed state against explicit recovery criteria
  • Track progress against recovery criteria as new information arrives
  • Surface residual risk and incomplete recovery before closure
After authorized failover Recovery verification
Stabilization in progress

Did it actually work?

Current operational state
  • Critical workloads transferred to Dallas
  • Service stabilized
  • Data Hall A thermal load reduced after workload transfer; local cooling recovery continues
Data Hall A representative equipment inlet temperature
Data Hall A representative equipment inlet temperature recovery Data Hall A representative equipment inlet temperature rises from approximately 23°C before degradation to above 30°C, then decreases after failover at minute 45 to 27.2°C at minute 75. The defined operating range is 18–27°C. Defined range (18–27°C) 32 30 27 24 21 18 Failover initiated Current 27.2°C -10 0 15 30 45 60 75 Minutes from degradation onset Temperature (°C)

Failover began 45 minutes after degradation started. After another 30 minutes, inlet temperature is decreasing but remains above the defined 18–27°C operating range.

Recovery criteria vs. observed
Recovery Criterion Observed Status
Critical workloads in Dallas Stable
Replication state Verified
Data Hall A representative equipment inlet temperature
27.2°C, decreasing
 Above defined range
Defined operating range
18–27°C
 Not yet within defined range
Equipment protection measures Active
Stability window
Not started
Temperature remains above the defined range
 Criterion not yet met
Not yet ready for closure: Temperature remains above the defined operating range. The 30‑minute stability period has not started, and local cooling recovery is still in progress.
Next verification: 5 min
05 Learning

What should the organization learn?

Turns validated outcomes into reusable operational experience while preserving findings, evidence and follow‑up.

  • Preserves validated findings, evidence and issues that still require follow‑up
  • Builds the event record from work already performed
  • Makes relevant experience available when similar situations occur again
Validated event record Critical Cooling Loss, Data Hall A
Event closed • Root cause confirmed

Data Hall A returned to its operating range and completed the required 30‑minute stability period. Recovery was verified, and the subsequent investigation established the root cause.

Root cause confirmed

Improperly resealed refrigerant service connection after CU‑04 compressor service caused a slow refrigerant leak and progressive loss of cooling capacity.

Validated findings
1

Higher compute density contributed to a shorter thermal response window

Higher compute density accelerated thermal escalation, reducing the available response time compared with the historical incident.

2

Post‑maintenance acceptance evidence was incomplete

The maintenance record lacked verification of refrigerant‑circuit integrity and delivered cooling performance before return to service.

3

Shared redundancy did not eliminate the common thermal exposure

Local redundancy shared Data Hall A’s thermal exposure and could not provide an independent recovery environment.

Investigation record
Critical Cooling Loss, Data Hall A Status: Approved
Evidence basis
  • Engineering leak test confirmed refrigerant loss at the service connection
  • Post‑repair recharge and recommissioning restored expected CU‑04 cooling performance
  • Current load records and thermal telemetry showed materially higher heat load and a shorter thermal response window than the historical comparison
View investigation report
Event status and follow‑up
Operational event status
• Recovery criteria satisfied
• Operational event closed after recovery validation
Control gap identified
Post‑maintenance verification requirements were insufficiently documented.
Procedure update proposed
Add refrigerant‑circuit integrity and cooling‑performance checks to post‑maintenance acceptance criteria.
Scheduled for engineering review
Thermal response basis
Reassess thermal‑response assumptions for higher‑density compute loads.
Future applicability
Apply these findings when evaluating cooling degradation following compressor maintenance.
Historical Experience

Validated findings, decisions, outcomes and follow‑up are preserved as Historical Experience and surfaced when relevant to future situations.

Validate Veras against a real operational challenge

Start with one operational challenge using the information and expertise your organization already has. No real-time integration is required.

What you bring

  • A relevant operational challenge
  • Available data, documentation and history
  • One or two operational specialists

What we explore together

  • Apply Veras to your operational context
  • Evaluate its understanding and decision support
  • Identify opportunities for improvement
Book a 15-min intro

Prefer email?
hello@getveras.com