Market Positioning SkoutLab occupies a unique position in the analytics ecosystem. While most tools focus on making humans faster at analysis , SkoutLab focuses on having AI do the analysis .
Human-Driven AI-Driven │ │ ┌────────────────────┼──────────────────────────────┼────────────────────┐ │ │ │ │ │ Traditional BI │ Augmented Analytics │ SkoutLab │ │ (Tableau, Power │ (Tellius, ThoughtSpot) │ (Autonomous │ │ BI, Looker) │ │ Analytics) │ │ │ │ │ │ "Build charts" │ "Ask questions faster" │ "AI finds │ │ │ │ insights" │ └────────────────────┴──────────────────────────────┴────────────────────┘
The Key Differentiator Aspect Tellius / ThoughtSpot SkoutLab Who asks questions Human AI proactively discovers Who finds answers AI assists human AI finds & validates Who interprets Human AI writes narrative, human reviews Core value ”More people can query data" "Get analysis without analysts” Replaces DA’s “data pull” work DA’s “analysis + reporting” work
Software Engineering Analogy:
Tellius = GitHub Copilot (helps you code faster)SkoutLab = AI Code Agent (understands the problem, writes the solution) Competitor Architecture Breakdown 1. Augmented Analytics (Tellius, ThoughtSpot) What they do: Natural language query + auto-insights + self-service BI┌─────────────────────────────────────────────────────────────────┐ │ User Interface │ │ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │ │ │ NLQ Input │ │ Drag-Drop │ │ Click │ │ │ │ "Show sales"│ │ Dashboard │ │ Explore │ │ │ └─────────────┘ └─────────────┘ └─────────────┘ │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ LLM Layer (Thin Shell) │ │ • Intent recognition: "show sales" → QUERY intent │ │ • Entity extraction: "last quarter" → time_range │ │ • Result explanation: Data → Natural language summary │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Semantic Layer (Core) │ │ Business Views: │ │ "Revenue" = SUM(orders.amount) WHERE status='paid' │ │ "Customer" = users JOIN subscriptions │ │ Permissions: User A sees only Region='US' │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Query Engine │ │ ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │ │ │ Text-to-SQL │ │ Query │ │ Cache │ │ │ │ (Templates) │ │ Optimization │ │ Layer │ │ │ └──────────────┘ └──────────────┘ └──────────────┘ │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Analytics Engine │ │ ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │ │ │ Anomaly │ │ Driver │ │ AutoML │ │ │ │ Detection │ │ Decomposition│ │ (Prediction) │ │ │ │ (Rules) │ │ (SHAP/Trees) │ │ │ │ │ └──────────────┘ └──────────────┘ └──────────────┘ │ │ │ │ ⚠️ This layer is DETERMINISTIC — not LLM-driven │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Data Connectors │ │ Snowflake │ BigQuery │ Redshift │ Postgres │ CSV │ └─────────────────────────────────────────────────────────────────┘
Key Limitations:
Reactive (waits for user questions)
Pattern matching, not true hypothesis testing
Outputs charts + bullets, not narrative reports
Requires semantic layer setup (weeks of config)
2. Palantir (Foundry / AIP) What they do: Enterprise data integration + ontology + AI applications┌─────────────────────────────────────────────────────────────────┐ │ Application Layer │ │ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │ │ │ Operational │ │ AIP │ │ Custom │ │ │ │ Dashboards │ │ (AI Chat) │ │ Apps │ │ │ └─────────────┘ └─────────────┘ └─────────────┘ │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ AIP Layer │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ LLM + RAG over Ontology │ │ │ │ • Chat interface to ask questions about data │ │ │ │ • Generates actions/workflows │ │ │ │ • Constrained by ontology definitions │ │ │ └─────────────────────────────────────────────────────────┘ │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Ontology Layer (Core Moat) │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ Object Types: Customer, Order, Product, Transaction │ │ │ │ Properties: customer.lifetime_value, order.margin │ │ │ │ Actions: approve_order(), flag_fraud(), assign_rep() │ │ │ │ Relationships: Customer → Orders → Products │ │ │ │ │ │ │ │ ⚠️ Requires months of professional services to build │ │ │ └─────────────────────────────────────────────────────────┘ │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Pipeline Layer │ │ ┌──────────────┐ ┌──────────────┐ ┌──────────────┐ │ │ │ Data │ │ Transform │ │ Data Quality │ │ │ │ Integration │ │ (Code/Low- │ │ Monitoring │ │ │ │ (100+ src) │ │ Code) │ │ │ │ │ └──────────────┘ └──────────────┘ └──────────────┘ │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Data Sources │ │ Enterprise Systems │ IoT │ External Data │ Files │ └─────────────────────────────────────────────────────────────────┘
Key Characteristics:
Strength: Deep enterprise integration, operational workflows, government contractsWeakness: Extremely expensive ($1M+ deals), 6-12 month implementationsTarget: Large enterprises with complex operational needsAI Approach: LLM as chat interface over pre-built ontology (constrained)
SkoutLab vs Palantir: Aspect Palantir SkoutLab Setup time 6-12 months Hours Cost $1M+ / year Self-serve pricing Target F500 operations Growth companies analytics AI role Chat over ontology Autonomous analysis Focus Operational actions Analytical insights
3. Databricks (Lakehouse + AI) What they do: Unified data platform + notebooks + ML/AI┌─────────────────────────────────────────────────────────────────┐ │ Application Layer │ │ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │ │ │ SQL Editor │ │ Notebooks │ │ Dashboards │ │ │ │ │ │ (Python/R) │ │ (Lakeview) │ │ │ └─────────────┘ └─────────────┘ └─────────────┘ │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ AI/ML Layer │ │ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │ │ │ MLflow │ │ Feature │ │ Model │ │ │ │ (Tracking) │ │ Store │ │ Serving │ │ │ └─────────────┘ └─────────────┘ └─────────────┘ │ │ │ │ ┌─────────────────────────────────────────────────────────┐ │ │ │ Mosaic AI (New) │ │ │ │ • AI/BI Dashboards (Genie) │ │ │ │ • Text-to-SQL │ │ │ │ • Agent Framework │ │ │ └─────────────────────────────────────────────────────────┘ │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Unity Catalog (Governance) │ │ • Data discovery │ │ • Access control │ │ • Lineage tracking │ │ • AI/BI semantic model │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Delta Lake (Storage) │ │ • ACID transactions │ │ • Schema evolution │ │ • Time travel │ └─────────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────────┐ │ Compute Layer │ │ Photon (SQL) │ Spark (ETL) │ GPU Clusters (ML) │ └─────────────────────────────────────────────────────────────────┘
Key Characteristics:
Strength: Best-in-class for data engineering + ML workflowsTarget User: Data engineers, data scientists, ML engineersAI Approach: Tooling for building AI (not AI that does analysis)Genie (AI/BI): New product, essentially ThoughtSpot-like NLQ
SkoutLab vs Databricks: Aspect Databricks SkoutLab Primary user Data engineers/DS Business analysts/Ops Value prop ”Build data + ML pipelines" "Get insights without building” AI approach Tools to build AI AI that does analysis Setup Complex (warehouses, compute, etc.) Upload CSV and go Output Notebooks, models, tables Analysis reports
Full Competitive Matrix Capability Traditional BI Augmented Analytics Palantir Databricks SkoutLab Setup time Weeks Days-Weeks Months Weeks Hours Who drives Human Human (AI assists) Human (AI assists) Human AI (human reviews) Hypothesis generation Manual None None Manual Automatic Statistical validation Manual Pattern matching Limited Manual code Automatic Narrative reports Manual Bullets Custom apps Manual Automatic Target user Analysts Business users Ops teams Data teams Business + Ops Price point $$$ $$$$ $$$$$ $$$$ $$
SkoutLab’s Moat Why Can’t Competitors Just Add LLM?
Architecture Matters Tellius/ThoughtSpot are built for reactive queries , not proactive exploration . Adding LLM is a thin shell, not a new capability.
Workflow Difference Our workflow is fundamentally different: AI generates hypotheses → AI validates → AI writes report . This requires ground-up architecture.
Statistical Rigor We do proper hypothesis testing with p-values and confidence intervals. Competitors do pattern matching and correlation.
No Semantic Layer Required Competitors need weeks of configuration. We work with raw data directly, making us accessible to companies without data teams.
Time Window Competitors are moving toward “agentic” capabilities. Our defensible moat needs to be built through:
Vertical Depth: Specialize in specific use cases (SaaS metrics, e-commerce, etc.)Statistical Rigor: True hypothesis testing, not just pattern matchingSpeed to Market: Build customer base and case studies before competitors catch up
Positioning Summary One-liner: “SkoutLab is the AI analyst that does the work, so your team can focus on decisions.”For different audiences: Audience Positioning Business buyer ”Get the insights of a data team without hiring one” Data team ”Automate routine analysis, focus on strategic work” VC/Investor ”Autonomous analytics — AI agents that replace analyst workflows”
Key insight: We’re not selling a “better BI tool” — we’re selling analysis capability . The buyer is someone who needs insights but doesn’t have (or can’t scale) analyst resources.
