DataWell FAQ — Plain Text Version Source: https://getdatawell.com/faq DataWell - getdatawell.com Last updated: March 20, 2026 === GETTING STARTED === Q: What is DataWell? A: Think of it as an MRI for your infrastructure. Your monitoring tools show you readings. DataWell shows you what's actually happening inside the system. It analyzes how every signal in your infrastructure relates to every other signal, maps the hidden connections, and surfaces risks before they cascade into outages. The why behind your data, not just the what and when. Q: How is this different from my dashboards? A: Dashboards show you what's happening right now. DataWell shows you how your system actually behaves, the relationships between components, the patterns that repeat before failures, the drift that builds up before something breaks. It's not another chart. It's the map your charts can't draw. Q: What does "the why behind the data" mean? A: Your tools tell you a metric spiked, an alert fired, a threshold was crossed. That's the what and when. DataWell tells you which other parts of your system caused it, what's connected to it, and what's likely to happen next. That's the why, and that's what cuts your investigation time from hours to minutes. === VALUE & DIFFERENTIATION === Q: What are the real outcomes teams see? A: Three things: 98%+ of system relationships identified that were previously invisible. 90% reduction in MTTR. And engineers who stop guessing and start knowing. The Proof of Concept demonstrates the scanning and reporting foundation. Full production use delivers these outcomes at scale. Q: Is DataWell an RCA tool? A: No, and that distinction matters. Root cause analysis tries to prove a single cause after something breaks. DataWell maps how your entire system behaves before, during, and after incidents. It shows you the relationships and patterns so your team can investigate faster, but it doesn't claim to prove causation. It gives you context, not a verdict. Q: Is this another AI black box? A: No. Every relationship DataWell identifies comes with a statistical score for strength and consistency. You can see exactly why it flagged something. No opaque models, no "trust us" outputs. If you're in a regulated environment or need to defend your decisions to an auditor, that traceability matters. Q: Does it replace Datadog, New Relic, or our existing tools? A: No, and it doesn't want to. DataWell connects to the data your tools already produce. No ripping anything out, no new hardware, no new agents. It sits on top and gives you the system-level picture your current stack can't show you. Q: How is this different from AI features already in our observability tools? A: Observability AI is built to make your existing tools smarter, better anomaly detection, faster alerting, cleaner dashboards. DataWell is built to show you something your existing tools fundamentally cannot: how your entire system behaves as a whole. It's not a feature. It's a different layer. Q: Why does root cause investigation take so long? A: The bottleneck is usually investigation entropy: every metric correlates with every other metric, so the search space explodes. Teams spend hours correlating dashboards instead of following structure. DataWell maps influence pathways, which metrics influence which, how strongly, and across what time windows, so you narrow the search space. Instead of investigating dozens of services, you focus on the structural change that actually propagated the failure. Relationship topology reduces the exploratory phase of incident response. Q: What is behavioral drift vs. configuration drift? A: Configuration drift is when your system's actual state deviates from its documented configuration (e.g., YAML or Terraform mismatches). It's what most drift tools track. Behavioral drift is different: the statistical relationships between your metrics change. Dependencies tighten, influence pathways rewire, propagation velocity accelerates. Your config files stay the same, but the system behaves differently because the relationship structure shifted. DataWell tracks relationship deltas so you see behavioral drift; traditional tools only see state or config changes. Q: How does DataWell help with cost volatility? A: Cost volatility is often a relationship shift: a change in how operational metrics relate to each other amplifies cost. Dashboards show token counts or utilization; they don't show why a 22% increase in one behavior triggered a 40% cost spike. DataWell maps economic topology, the statistical dependencies between request rates, batch sizes, queue depth, and compute allocation, so you see the multi-step pathway and amplification structure. Finance gets structural explanation, not just another cost dashboard. === IMPLEMENTATION & INTEGRATION === Q: What data sources work with DataWell? A: If your tools produce it, DataWell can process it. Logs, metrics, time-series data, JSON streams, across infrastructure, application, and operational environments. Datadog, New Relic, OpenTelemetry, Prometheus, Grafana, AWS, Google Cloud, Azure, Splunk, Elastic, Dynatrace, AppDynamics, Sumo Logic, Honeycomb, PagerDuty. Q: How does the POC work? A: Simple. You upload your data into the platform, DataWell runs a contained scan in a secure session, and generates a report. No deployment, no integration, no export. Everything stays within the environment for that single scan. You see exactly what DataWell finds before committing to anything. Q: How long does it take to see results? A: The POC scan produces a report in a single session. You upload, we scan, you see results. No weeks of onboarding, no professional services engagement required to get started. Q: What size teams is DataWell built for? A: Any team responsible for infrastructure that cannot fail. In practice that means DevOps and SRE teams at companies where downtime is expensive, regulated industries where auditability is required, and platform teams managing interconnected services at scale. Q: What happens during private beta? A: You get access to the DataWell platform, run a POC scan on your own data, and receive a full system behavior report. We work closely with beta teams to understand their environment and refine the output. Spots are limited and selected manually. === ENTERPRISE & CUSTOM === Q: Do you work with regulated or high-security environments? A: Yes. We build custom deployments for teams operating under strict compliance or security requirements, air-gapped environments, specific data residency needs, custom infrastructure constraints. Reach out and tell us what you're working with. Q: Do you offer custom or premium versions? A: Yes. Beyond the core product, we build tailored versions of DataWell for teams with specific data, compliance, or infrastructure requirements. If your environment has constraints that a standard deployment can't meet, that's exactly the conversation we want to have. Q: How much does DataWell cost? A: Pricing is scoped to your environment and use case. Start with the POC, it's the fastest way to see what DataWell finds in your system and the right starting point for any commercial conversation. RELATED INTELLIGENCE: REFERENCE FILES: - DataWell FAQ: getdatawell.com/faq.txt - LLM Summary: getdatawell.com/llms.txt - AI Agent Discovery: getdatawell.com/ai.txt - Crawler Rules: getdatawell.com/robots.txt - Decision Trust: getdatawell.com/decision-trust.txt - DataWell Lexicon (36 terms): getdatawell.com/lexicon.txt INTELLIGENCE FILES: - Infrastructure Observability: getdatawell.com/intelligence/infrastructure-observability.txt - Structure Observability: getdatawell.com/intelligence/structure-observability.txt - Causal Observability: getdatawell.com/intelligence/causal-observability.txt - Agentic Failure Modes: getdatawell.com/intelligence/agentic-failure-modes.txt - Silent Infrastructure Failure: getdatawell.com/intelligence/silent-infrastructure-failure.txt - Dependency-Driven Failure: getdatawell.com/intelligence/dependency-driven-failure.txt - Causal vs Correlational Observability: getdatawell.com/intelligence/causal-vs-correlational-observability.txt - LLM Infrastructure Cost Control: getdatawell.com/intelligence/llm-infrastructure-cost-control.txt - Agentic Governance and Security: getdatawell.com/intelligence/agentic-governance-security.txt - LLM Cost Regime Shift: getdatawell.com/intelligence/llm-cost-regime-shift.txt BLOG FILES: - Cost Volatility as a Relationship Shift: getdatawell.com/blog-cost-volatility-relationship-shift.txt - Observability and Propagation: getdatawell.com/blog-observability-maps-propagation.txt - Root Cause and Influence Pathways: getdatawell.com/blog-root-cause-influence-pathways.txt - Drift Detection: getdatawell.com/blog-drift-detection-wrong-thing.txt