FAQ

Get answers, fast

Everything you need to know about Skyriq & OscillAI

Explore our most common questions about optimizing storage across multiple clouds through Skyriq and OscillAI. We’re here to help you navigate, simplify, and succeed.

General

Why is cloud storage so expensive?

Cloud storage is priced by provisioned capacity, not actual usage. Engineers create 500GB volumes for applications that typically use 150GB, creating massive waste:

Over-provisioning for peak: Sized for worst-case to avoid downtime
Orphaned volumes: Forgotten after application migration
Developer guesswork: 55% of developers estimate storage needs without data
No real-time optimization: Quarterly reviews miss dynamic changes

Result: 30-45% of storage budgets disappear into unused capacity.

Which teams benefit most from OscillAI?

OscillAI is valuable across organizations, with highest ROI for:

DevOps/Infrastructure Teams:

Manage thousands of volumes across environments
Currently spend 10-20% of time on manual capacity planning
Benefit: Eliminate manual resizing tickets; focus on architecture

FinOps/Cloud Finance Teams:

Track cloud waste but lack automation to fix it
Currently only see quarterly cost reviews
Benefit: Real-time cost control; automated invoice reduction

Engineering Leaders (CTOs/VPEs):

Protect uptime; currently over-provision storage for safety
Balance innovation velocity against cloud cost creep
Benefit: Safe automation prevents disk-full incidents

CFOs/Finance:

Demand cost predictability; cloud storage is unpredictable
Need to show margin improvement without slowing growth
Benefit: Recurring monthly savings; improved unit economics

How does OscillAI differ from manual optimization approaches?

Manual reviews are reactive and labor-intensive; OscillAI is continuous and autonomous.

Manual reviews happen quarterly or annually due to scheduling and resource constraints. By the time a quarterly review occurs, months of waste have already accumulated. People must allocate time to capacity planning, analysis, and implementation of changes. The process is inherently slow—weeks typically pass between decision and action.

OscillAI operates continuously in real-time. Changes happen immediately when conditions warrant, not months later. The system uses live, current metrics instead of historical snapshots that may no longer reflect reality. The infrastructure overhead is minimal—2% rather than 10-20% of engineering time. The approach is completely comprehensive, reviewing and optimizing every volume, not just the obvious targets.

Result: Manual review captures 10-20% cost reduction from the lowest-hanging fruit. OscillAI achieves 50-70% reduction through continuous, comprehensive optimization that humans cannot match.

How does OscillAI compare to Native cloud scaling features?

Native AWS tools have architectural limitations that OscillAI overcomes.

Scaling down: AWS Auto Scaling cannot scale volumes down because reduced EBS = lost AWS revenue. This is a fundamental business conflict for cloud providers. OscillAI has no such conflict—smaller volumes benefit both you and us.

Speed: AWS operates on a 6-hour minimum window due to architectural constraints. OscillAI scales in minutes or seconds as needed.

Autonomy: AWS recommendations are advisory only—you must still manually implement changes. OscillAI executes autonomously within your guardrails.

Predictive capability: AWS uses basic rules for scaling decisions. OscillAI uses AI-driven forecasting that predicts demand 24-48 hours ahead.

Cost model: AWS includes scaling in the CSP margin (less incentive to downsize). OscillAI exists purely to optimize your costs.

Dashboard: AWS metrics live in CloudWatch. OscillAI provides a dedicated dashboard optimized for storage visibility and control.

The fundamental difference: AWS native scaling is a compliance feature for their platform. OscillAI is a purpose-built optimization engine designed specifically for block storage cost reduction.

‍

What's a typical real-world implementation story?

Company profile: Mid-market technology company, 3 AWS regions, 750TB block storage provisioned.

Before OscillAI:

Actual utilization was only 280TB (37% of provisioned). Monthly waste was $23,625 in unused capacity. One DevOps engineer spent 5-8 hours per week on capacity reviews and adjustment. Each region was optimized independently without coordination, leading to inconsistent policies.

Implementation timeline:

Week 1 - 2: Capacity audit showed optimization opportunity of $283,500 annually. Guardrails were configured to match existing operational policies. Deployment was completed in 45 minutes across all three regions.

Month 1 - 2: Automatic consolidation and optimization began. 60% cost reduction was achieved through a combination of immediate rightsizing and predictive scaling. New monthly cost dropped to $15,000. Net savings of $22,500/month started immediately.

Results:

Payback period was 4 days from deployment. Year 1 ROI was 670x (solution cost was minimal, annual savings were $270,000). Implementation time totaled 6 hours across all engineering and operations. The freed-up DevOps capacity enabled the team to focus on multi-region architecture improvements.

‍

Technical Capabilities

What data does OscillAI access?

OscillAI accesses only metadata and metrics—never your actual data:

✅ Provisioned capacity (GB allocated)
✅ Utilized capacity (GB actually used)
✅ I/O operations per second (IOPS consumed)
✅ Provisioning timeline (growth patterns)
❌ No access to encrypted data, files, or content
❌ No access to application logic or secrets
❌ No data leaves your cloud account

Compliance:

Works with HIPAA, GDPR, SOC2, PCI-DSS environments
No data residency concerns; runs entirely in your account
Operates as read-only observer with execution permissions only for storage operations

‍

How does OscillAI prevent disk-full incidents?

Multi-layered approach ensures volumes never run out of space unexpectedly.

Real-time monitoring: OscillAI continuously watches utilization trends and alerts when approaching configured thresholds (e.g., 80% full). The system triggers expansion before reaching critical capacity, not in response to an emergency.

Predictive scaling: Using demand forecasting, OscillAI predicts spikes 24-48 hours in advance. It pre-scales before you need the capacity, eliminating cold-start delays that occur with reactive approaches.

Safety margins: Default guardrails maintain a configured headroom (e.g., always keep 10% free space). These margins are customizable based on your workload characteristics.

Audit trail: Every scaling action is logged with timestamp, reason, previous utilization, and growth trajectory. You see exactly why each decision was made.

Result: Disk-full incidents become preventable—not reactive emergencies requiring emergency support calls at 3am.

‍

How does OscillAI handle Kubernetes persistent volumes?

Full compatibility with Kubernetes persistent volumes across AWS EBS.

Supported workloads:

Kubernetes persistent volumes (PVs) backed by AWS EBS in any configuration. Container Storage Interface (CSI) drivers for dynamic provisioning. StatefulSet workloads requiring stable, long-lived storage. Multi-AZ and region deployments with volume replication.

How it works:

OscillAI monitors and optimizes the underlying cloud volumes (EBS layer), while Kubernetes sees stable performance. The optimization is completely transparent to Kubernetes—no configuration changes required. Kubernetes continues managing containers; OscillAI manages the storage layer.

Key benefit:

Kubernetes deployments are highly portable—they can move between clusters without losing optimization. The storage optimization travels with your volumes automatically.

‍

What are "guardrails" and how do they protect my workloads?

Guardrails are policy limits you set to control OscillAI's optimization:

Available guardrails:

Minimum volume size: Never scale below (e.g., 50GB minimum for database)
Maximum scaling rate: Limit expansions per day (e.g., 10% daily max)
Do-not-optimize volumes: Exclude critical volumes from automatic scaling
Scaling windows: Restrict changes to maintenance windows
IOPS thresholds: Trigger scaling only if IOPS exceed X
Audit requirements: All actions logged for compliance review

How they work:

100% customer-controlled; no "magic" in background
Audit trail: Every expansion/contraction logged with timestamp, reason, approval
Manual override: Teams can adjust limits in real time if conditions change

‍

What's the difference between volume scaling and workload optimization?

Two complementary approaches to cost reduction:

Volume Scaling (OscillAI's focus):

Adjust provisioned capacity UP and DOWN based on actual usage
Example: 1000GB EBS → 300GB actual usage = scale to 400GB (with headroom)
Benefit: 60-70% cost reduction on individual volumes

Workload Optimization (Complementary):

Consolidate or redistribute workloads to fewer, larger volumes
Implement compression, deduplication, tiering
Example: Migrate cold data to cheaper archive tiers

OscillAI strength: Automates volume scaling (hard without AI), complements human-led workload optimization.

‍

Integration and Deployment

How do we manage OscillAI across multiple AWS accounts

OscillAI supports multi-account AWS environments with flexible deployment approaches.

Per-account agents (Recommended for most):

Deploy an independent agent in each AWS account. Each agent manages volumes within its own account. Unified dashboard shows all accounts with cost attribution. Clear separation of concerns and billing. Simplest operational model for most organizations.

Multi-account management:

Single dashboard with account-level cost attribution. Consistent policies across accounts or per-account customization. Centralized audit logging and compliance reporting. Exportable compliance reports for regulatory review.

‍

What happens during AWS maintenance or updates?

OscillAI respects AWS platform events and maintains service during maintenance windows.

During AWS maintenance:

OscillAI continues monitoring (read-only access is typically unaffected). Expansion operations pause automatically if EC2 API is unavailable. Operations resume immediately when AWS services restore. No manual intervention required.

Your options:

Automated (default): OscillAI pauses operations during known AWS maintenance windows. Custom windows: Define specific maintenance windows when scaling is disabled. Full control: Manually pause OscillAI anytime via dashboard or API.

Communication:

You're notified of AWS planned maintenance. OscillAI automatically detects when API availability is restored. Operations resume without requiring any action from you.

‍

How does OscillAI handle multi-region deployments?

Native support for distributed architectures across AWS regions.

‍Optimization approach:

Each region's volumes are treated independently (region-specific optimization patterns). Replication constraints are respected (mirror volumes scale together). Unified dashboard shows all regions. Cost attribution by region, availability zone, or resource tag.

Example - 3-region setup:

Region 1 (US-East) manages 200TB optimized independently. Region 2 (US-West) manages 150TB optimized independently. Region 3 (EU) manages 100TB optimized independently. Single dashboard provides full visibility and control across all regions. One set of controls for consistent governance.

‍

What if I want to pause or disable OscillAI temporarily?

Complete control to pause, disable, or re-enable OscillAI at any time.

Temporary pause:

Pause optimization via dashboard for scheduled maintenance or testing. All existing volume configurations preserved. Resume with one click when ready. Useful for planned capacity tests or temporary workload changes.

Disable completely:

Disable OscillAI without uninstalling the agent. Configuration is preserved if you want to re-enable later. Volumes maintain their current size—no scaling occurs while disabled.

Re-enable anytime:

Re-enable through dashboard and optimization resumes immediately. No reconfiguration needed—previous guardrails and policies remain in place.

Manual override:

Override automatic scaling decisions for specific volumes temporarily. Useful when you know your workload is changing and want to manage scaling manually for a period.

‍

How does OscillAI handle snapshots and backup volumes?

OscillAI works seamlessly with AWS backup and snapshot infrastructure.

Snapshot behavior:

Snapshots capture the current volume state. OscillAI continues optimizing the source volume independently. Snapshot size reflects the source volume's current state at capture time. Restore operations create new volumes that OscillAI immediately begins optimizing.

Compliance consideration:

All operations are audited and logged. Optimization decisions are recorded for compliance verification if required.

‍

Cloud Roadmap and Support

When will OscillAI support other cloud providers?

Multi-cloud expansion is core to our product roadmap.

Current status:

AWS (All EBS types) - Production-ready and available now. Microsoft Azure - Available April 2026. Google Cloud Platform - Coming later in 2026.

Why phased rollout:

Each cloud platform has unique optimization characteristics requiring engineering effort. AWS launch allows us to perfect the product before scaling. Early adopters gain 6+ months of proven optimization experience before other clouds launch. Phased approach ensures quality and stability at each stage rather than rushing unsupported platforms.

Multi-cloud customers:

Start optimizing AWS immediately. Azure optimization joins April 2026 seamlessly. Google Cloud optimization added later 2026. Unified dashboard will eventually manage all clouds from one interface. No re-learning—same concepts work across clouds.

Can we use OscillAI across both AWS and Azure?

Yes, once Azure support launches in April 2026, multi-cloud optimization becomes seamless.

Timeline:

Now: Optimize AWS volumes. April 2026: Add Azure optimization. Later 2026: Add GCP optimization if needed.

How it works:

Single dashboard manages optimization across clouds. Consistent guardrails and policies (or per-cloud customization). Cost attribution by cloud, account, or resource tag. Unified billing and reporting.

Setup:

Provide credentials for each cloud. OscillAI handles the rest. No re-training—same concepts work across clouds. Migration between clouds doesn't disrupt optimization.

‍

What if we're using a cloud provider OscillAI doesn't support yet?

Multiple options available while we expand support.

Option 1: Start with supported cloud

Deploy OscillAI on AWS today. Begin capturing 50-70% savings immediately. Seamlessly add other clouds when support launches (April 2026 for Azure, later 2026 for GCP).

Option 2: Early access program

Contact sales for beta access to upcoming cloud platforms. Help us refine support for your platform. Get priority access at official launch.

Option 3: Contact sales

Discuss your specific cloud setup. Explore timeline for your platform. Determine best approach for your company. Custom arrangements may be available for strategic customers.

‍

What kind of customer support does Skyriq provide?

Comprehensive support designed to ensure your success with OscillAI.

Included support:

Email support for general questions and troubleshooting. Dashboard and documentation access. Regular product updates and improvements. Quarterly business reviews (for enterprise customers).

Onboarding support:

Dedicated support during implementation (weeks 1-4). Configuration guidance and best practices. Initial optimization review and tuning.

Escalation path:

Technical issues escalated to engineering team. Priority support for production issues. Custom support arrangements available for enterprise customers.

Community:

Access to user community for peer support. Feature requests and feedback channels. Best practices and case studies from other customers.

‍

What's the difference between AWS and Azure/GCP optimization?

Core optimization logic is the same; platform specifics differ:

Common across all clouds:

Real-time scaling based on utilization
Predictive forecasting for demand spikes
Autonomous volume management
Guardrails and customer control

Cloud-specific differences:

AWS EBS: Volume types (gp3, io1, st1, sc1), snapshot lifecycle
Azure Managed Disks (April 2026): Disk types (Standard SSD, Premium, Ultra), tiering
GCP Persistent Disk (2026): Zonal vs. regional, machine type constraints

Optimization approach:

Each cloud's native APIs used to best effect
Consistent ROI (50-70% savings target)
Same dashboard experience across clouds

‍

Security & Compliance

How does OscillAI handle security?

Enterprise-grade security controls:

Data security:

Metadata only; no access to data content
Zero-knowledge design: You retain all encryption keys

Access control:

AWS IAM integration (role-based permissions)
Least-privilege design (only necessary permissions)
No shared access with other customers
Multi-account isolation supported

Audit & logging:

All actions logged for Audit
90-day retention minimum
Exportable audit reports for compliance

How does OscillAI ensure data residency?

Architecture:

OscillAI agent runs in your AWS account
No data aggregation with other customers
Metrics stay local: CloudWatch (your account)
No external data transfers for optimization

Data isolation:

❌ Volume names, tags, content never leave your account
❌ IOPS data, utilization patterns not sent externally
❌ No competitive intelligence harvesting
❌ No third-party data brokers

Verified security:

Runs on AWS infrastructure (your region)
Third-party security audit available on request
Audit trail your auditors can inspect

‍

Who owns the optimization decisions?

Your control mechanisms:

Guardrails: You set policies; OscillAI operates within them
Audit: You review all actions; approve or reject changes
Disable: You can pause optimization at any time
Workload exclusion: You designate volumes as "do not optimize"
Scaling windows: You restrict changes to maintenance windows

How it works:

You define guardrails → OscillAI optimizes autonomously within bounds
NOT: OscillAI decides → you react to changes

Typical governance flow:

IT/DevOps team sets guardrails (volume size limits, scaling rates)
FinOps team reviews cost attribution and approves optimization policy
Security team reviews audit logs monthly
OscillAI operates autonomously within all constraints

‍

How do we get started with OscillAI?

Simple 3-step onboarding process:

Step 1: Free Evaluation (No credit card required)

Contact us for sign up at www.skyriq.com
Share read-only AWS credentials
Receive 1-week capacity audit (no installation)
Output: Savings projection, waste breakdown, optimization roadmap

Step 2: Proof of Concept (1-2 weeks)

Deploy agent to non-production environment
Set initial guardrails
Monitor optimization in sandbox
Validation: See real-time cost reduction on test volumes

Step 3: Production Deployment (1 week)

Deploy to production with agreed guardrails
Enable full optimization
Begin cost reduction on live volumes
First results: Savings visible on next AWS bill

Glossary

What are the terms of reference?

Autonomous Scaling: Automatic volume expansion/contraction based on real usage without human intervention.

Block Storage: Low-level storage accessed in fixed-size blocks; used for databases, VMs, containers.

Buffer: Extra capacity maintained as safety margin (e.g., "always keep 10% free").

CloudWatch: AWS native monitoring service; OscillAI integrates with CloudWatch metrics.

Guardrails: Policy limits controlling how aggressively OscillAI optimizes (you define these).

IOPS: Input/Output Operations Per Second; measure of storage performance.

Over-provisioning: Allocating more storage than actually used, for safety.

Provisioned capacity: Amount of storage you allocate (not necessarily used).

ROI: Return on Investment; measure of cost savings vs. solution cost.

Scaling down: Reducing provisioned capacity (unique to OscillAI; CSPs can't do this easily).

Scaling up: Increasing provisioned capacity (all storage solutions can do this).

Thin Provisioning: Allocating storage dynamically as needed, not all upfront.

Utilization: Percentage of provisioned capacity actually used by application.

Everything you need to know about Skyriq & OscillAI

General

Technical Capabilities

Integration and Deployment

Cloud Roadmap and Support

Security & Compliance

Glossary

Ready to simplify your cloud?