AI-Native
Investment Intelligence
WeChat + X/Twitter signal capture, AI deal evaluation, full-cycle pipeline management, and portfolio monitoring — all in one AI-powered platform.
Challenges Across the Investment Lifecycle
From signal discovery to deal execution to portfolio monitoring — friction compounds at every stage.
Information Overload
100+ articles daily from WeChat, RSS, and X. 90% is noise, but you can't afford to miss the 10% that matters.
Fragmented Channels
Valuable signals scattered across WeChat, Twitter, news sites, and databases. No unified view, constant context-switching.
Chaotic Pipeline
Deal stages tracked in spreadsheets and chat groups. No structured workflow, easy to lose track of progress and ownership.
Slow Decision Cycles
DD materials scattered, IC scheduling manual, review feedback siloed. Weeks lost on process, not analysis.
Portfolio Blind Spots
Quarterly data collected via email and spreadsheets. No real-time visibility into burn rate, runway, or risk signals.
Reactive, Not Proactive
Problems discovered in the next board meeting, not when they start. No automated alerts, no early warning system.
Signal Discovery
AI-powered signal capture and filtering. Find opportunities before they become obvious.
Multi-Source Monitoring
WeChat, RSS, and X/Twitter — continuously monitoring multiple channels. One setup, always listening.
Intelligent Extraction
AI extracts companies, founders, investors, funding rounds, and key events from every source.
Thesis-Aligned Classification
Signals classified as Target, Worth Reading, or Skip — aligned with your investment focus.
Evidence-Based Reasoning
No black boxes. Every recommendation comes with a transparent evidence chain you can verify.
Efficient Triage
Three-column inbox for rapid review. Process hundreds of signals in minutes, not hours.
Research-Ready Context
Export structured context to any AI assistant for deeper analysis with full background.
How It Works
Capture
Automated crawling of WeChat, RSS, and X/Twitter. Raw content transformed into structured intelligence.
Understand
AI evaluates every signal: relevance scoring, entity extraction, thesis alignment, evidence chain generation.
Act
Top signals surfaced in your inbox. One-click to add to pipeline, export context, or deep dive.
Data Sources
Deal Management
Structured pipeline for collaborative decision-making. From discovery to close, everything in one place.
Visual Pipeline
S0 Discovery → S1 Pre-DD → S2 DD → S3 IC → S4 Negotiation → S5 Portfolio. Clear stages, clear progress.
Rich Project Profiles
Company overview, team background, funding history, related signals — all consolidated in one view.
Collaborative Review
Multi-person scoring, comments, and tags. Team decisions with full audit trail.
Automated Weekly Reports
Auto-generated summaries of new projects, stage movements, and key activities each week.
Feishu Integration
Sync updates to Feishu automatically. Keep your team informed in the tools they already use.
One-Click Actions
Move stages, add notes, assign owners — streamlined workflows for fast-moving decisions.
Portfolio Management
From quarterly data collection to risk alerts — AI-powered post-investment monitoring for every portfolio company.
Quarterly Data & Trends
Track revenue, burn rate, runway, and key metrics across all portfolio companies. Spot trends before they become problems.
Automated Risk Alerts
AI monitors financial health indicators and triggers alerts when thresholds are breached. Never miss a warning sign.
AI-Powered Report Verification
Upload financial statements and let AI cross-check against reported numbers. Identify discrepancies instantly.
Resource Platform
Connect portfolio companies with partners, recruiters, and resources. Track usage and ROI across your ecosystem.
Feishu Bitable Sync
Bi-directional sync with Feishu Bitable. Edit on the web, changes flow back to your shared workspace automatically.
Role-Based Access
Partners see everything. Associates see their own companies. Granular permission control across your investment team.
Coming Next
Expanding coverage and deepening intelligence. Building in the open.
Signal Source Expansion
Extending our ears to where early signals emerge
arXiv & Research Papers
Daily monitoring of ML/AI papers. Link innovations to commercial opportunities.
- Daily paper digest
- Breakthrough detection
- Author tracking
- Citation network
GitHub & Open Source
Track trending repos, star velocity, and developer adoption signals.
- Trending repositories
- Star growth tracking
- Contributor analysis
- Ecosystem health
Technical Intelligence
Deep understanding for AI-focused investments
SOTA Analysis
Compare startup claims against state-of-the-art benchmarks. Verify technical credibility.
- Benchmark comparison
- Paper-to-product tracing
- Technical feasibility
- Innovation assessment
Portfolio Company Portal
Self-service portal for portfolio companies
Company Self-Service Portal
Portfolio companies log in to submit quarterly data, access resources, and view their own analytics.
- Self-service data submission
- Resource marketplace access
- Analytics dashboard
- Direct communication channel
System Architecture
Enterprise-grade AI architecture with clear separation of concerns.
Agent Layer
Dynamic orchestration for new scenarios
Workflow Layer
Deterministic, reproducible execution chains
Skill Layer
Composite capabilities combining tools and LLM
Tool Layer
Atomic, deterministic operations
Core Capabilities
Foundational principles
Security by Design
All AI actions go through Tools. LLM cannot access database directly or generate code.
Evidence-First
No black-box scores. Every recommendation shows the reasoning chain for verification.
RLHF-Driven
User actions continuously improve recommendations. Your preferences shape the AI.
Plugin Architecture
Data sources and renderers are pluggable. Add new integrations without touching core code.
Evaluation Center
Continuous quality assurance for AI-powered features.
Prompt Benchmarking
Systematically test and measure prompt effectiveness across different scenarios.
Regression Tests
Ensure AI quality doesn't degrade with automated regression testing suites.
Model Comparisons
A/B test different models and configurations to find the optimal setup.
Prompt Versioning
Track prompt changes over time with full version history and rollback.