Not Opinions. Structure. Built on Claude Code. Data as of April 11, 2026.
Today's Intelligence Feed April 11, 2026
262 stories ingested from 38 RSS feeds across US, China, EU, India. Consolidated, scored, and ranked by structural importance.
Salesforce Agentforce Generated 1M+ Recommendations Maintaining Data Integrity
Salesforce Engineering | Score: 18.0 | US | April 8, 2026
Revenue Orchestration team built Sales Agent using Agentforce. Cockpit-to-Vault movement. Enterprise AI agents in production, not demos.
Cockpit DepthAgent DynamicsUSScore: 18
EngineAI Raises $200M Series B, Valuation Exceeds RMB 10 Billion
Pandaily | Score: 15.0 | China | April 10, 2026
Chinese robotics startup raises $200M. Humanoid robots across industries. Capital flowing to physical AI infrastructure. Elo Score: 80 (highest this week).
Revenue QualityElo: 80China
Alibaba Qwen Captures Over 50% of Global Open-Source Downloads
SCMP Tech | Score: 14.0 | China | April 10, 2026
Qwen 3.5 dominates open-source. China's parallel stack is real. Protocol layer play via open-source model dominance.
Consequence LayerScore: 14China
SiFive Leads Week With $400M For Custom Chip Designs
Crunchbase News | Score: 12.0 | US | April 10, 2026
$400M into custom silicon. Capital flowing to compute infrastructure (Vault), not application layer (Studio).
Vault / InferenceScore: 12US
India: Bidso Bags Rs 63 Cr, Nava Raises $22M, Swiggy Reshuffle
StartupTalky + YourStory [2 sources] | Score: 12.0 | India | April 10, 2026
Multi-source convergence signal. Blume Ventures leading. RBI proposes UPI transfer delays. India ecosystem layer moves.
India EcosystemScore: 12India2 sources
Replit: How PMs Ship Faster Using Agentic Workflows
Replit Blog | Score: 11.0 | US | April 6, 2026
PMs using Replit agentic workflows. Agent-directed work replacing human-directed. X-axis movement on TCS map.
Agent DynamicsCockpit to VaultScore: 11
Alibaba Launches 10,000-Card Computing Cluster
SCMP Tech | Score: 10.0 | China | April 8, 2026
10,000-card intelligent compute cluster. China building its own Vault. The infrastructure race has three geographies now.
Vault / ComputeScore: 10China
Evaluator Verdicts: 27 signals scored. COUNCIL 24 KILL 3. No natural SURFACE this cycle. Top signal force-promoted for review.
Trend Analysis: What You Need to Know
Stories clustered into structural trends using Elo-inspired ranking. Not what happened. What it means.
Where Capital Flows
9 signals | Top Elo: 80 | Highest priority
Capital moved this week across India, China, US. The real question is not how much was raised. It is which layer the capital is targeting. Money flowing to compute infrastructure (Vault) has different structural implications than money flowing to AI app builders (Studio). Follow the layer, not the headline.
Key question: Is money flowing to capability (Studio) or consequence (Cockpit/Vault)?
The Infrastructure Race
4 signals | Top Elo: 66 | Accelerating across 3 geographies
Signals point to an accelerating infrastructure race across 3 geographies. SpaceX, NVIDIA, Palantir, Alibaba in the frame. The compute layer is the Vault. Capital flows to whoever controls inference costs. Second-order question: as inference gets cheaper, which application-layer companies benefit, and which lose pricing power?
Key question: Who is building the Vault?
Models Are Commoditizing
2 signals | Top Elo: 66 | Existential for Studio companies
Alibaba Qwen capturing 50%+ of open-source downloads. Every model release makes the creation layer cheaper. Great for users. Existential for companies whose only moat is model quality. Structural winners have consequence layers that survive model upgrades.
Key question: If the model advantage disappears tomorrow, what is left?
Top Stories by Elo Score
| Elo | Story | Type |
|---|
| 80 | EngineAI Raises $200M Series B | Funding |
| 75 | SpaceX posted nearly $5 billion loss in 2025 | Earnings |
| 66 | The race to control compute; Nava's growth plan | Funding |
| 56 | SiFive Leads With $400M For Custom Chips | Funding |
| 50 | Alibaba Qwen 50%+ global open-source downloads | Product |
Council Output: Cursor ($29B) vs Replit ($9B)
Real 6-seat debate from April 4, 2026. Phase 1 grounds the facts. Phase 2 argues from verified data. SD resolves.
THE BULL
Cursor at $29B is buying the most valuable real estate in enterprise software: the cognitive workflow of the developer.
60% enterprise signals Organisational and Relational layers forming. $2B ARR doubling in 3 months is expansion revenue inside accounts. Platform behaviour, not tool behaviour.
Key data: $1.2B enterprise revenue. Multi-year contracts. Real switching-cost architecture.
If Replit's runtime moat is structural, why is enterprise still individual users, not 60% enterprise revenue?
THE STRUCTURAL
Cursor: Cockpit. 1.5 layers confirmed (Habitual + partial Organisational). Aspiring Vault via agent micromodels.
Replit: Cockpit-to-Vault bridge. 2 confirmed layers (Technical + Habitual). Up to 4.5 with forming Talent and Protocol layers.
Prediction: Cursor is priced like a Vault company but is a Cockpit company. Replit is a Vault company priced like Cockpit. $9B on deeper layers = structural undervaluation.
Does Cursor's aggregate codebase context across millions of enterprise repos constitute a Protocol layer?
THE CONTRARIAN
Assumption challenged: code is still the primary interface for building software.
Both bull and structural assume the contest is for the best seat inside the coding workflow. What if the building is being demolished?
50M Replit users are not mostly engineers. They are trying to stop coding entirely. Lovable, v0, Bolt prove the specification layer is decoupling from execution. When complete, an AI code editor is a fax machine with a better UI.
If 60% of Cursor's enterprise revenue comes from companies actively eliminating developer headcount, what is the long-term demand curve?
COMMON WISDOM
The market believes Cursor is the defining winner of the AI coding wave. $29B on $2B ARR with 60% enterprise.
This consensus is narratively driven, not structurally grounded. It pattern-matches to "fastest-growing SaaS" without counting consequence layers. The market prices growth rate, not dependency depth.
SD's Decision Point: Is code the permanent construction medium for software, or an intermediate layer being abstracted away? The answer determines whether Cursor's $29B or Replit's $9B is the mispricing.