Project Review: Graphify — Self-learning company brain via knowledge graphs
Startup Profile
- Core Product: Knowledge graph generator/query engine for repos, docs, PDFs, SQL schemas, etc.
- Tech Stack/Model: Proprietary way(s) or toolset (Graphify .cmd) turning data into a kindest possible knowledge graph for Claude queries + LESSONS.md logging mechanism for self-learning reflection.
- Funding & Stage: Recently joined YC S26.
Growth Metrics certain as of launch window (April 5th): 73k stars, 2.2M downloads in ~2.5 months.
The Moat & Market Potential
Unlike standard RAG which may suffer from lost context and high token costs, Graphify creates a structured reasoner capable of using significantly fewer tokens per query (~71x less). Its ability to implement "self-learning" by tracking successful vs. failed answers through a persistent own memory (
LESSONS.md
) provides a layer of agency that simple vector search lacks. This positions it not just as a code search tool but as an intelligent enough agentic index meant to act as the live brain of a team's documentation and codebase.
Pros & Cons / Red Flags
- Strengths: High efficiency/low latency via reduced token usage; strong community momentum with fast download growth or organic traction으로 proofed ($
- Risks: Dependency on underlying LLMs like Claude for querying capabilities; potential complexity in maintaining graph integrity during frequent repo updates; risk of being part of any specialized niche if larger models (OpenAI/Anthropic) adopt better native long-context reasoning without needing external graphs.
Project Review: ORBIS — Real-time causal macroeconomic mosaic
Startup Profile
- Core Product: Causal monitoring platform using real-time data fragments (Macro, news, supply shocks).
- Tech Stack/Model: 26-node causal graph capable of recomputing asset exposure based on market pressure shifts.
- Funding & Stage: Subscription model at $49/mo.
The Moat & Market Potential
ORBIS targets the high-value ability to synthesize fragmented public datasets—such as FRED and live tape—into a cohesive way that traditional tools lack by manually processing certain connections too slowly. By automating the "mosaic" or assembly of these signals through a causal graph rather than simple correlation, it provides actionable intelligence regarding mispricing and systemic shock risks.
Pros & Cons / Red Flags
- Strengths: High transparency via cited ThesisCards; addresses timely enough needs in volatile markets where manual reassessment is impossible even for professionals.
- Risks: Highly specialized niche potentially limited if not able to scale node count significantly beyond current setup; relies heavily on quality input from external sources like global news and macro indicators which can be noisy.
Bottom Line: Graphify shows massive potential (YC backed) as an agentic infrastructure tool with extreme efficiency gains; ORBIS offers a sophisticated technical edge for real-time risk assessment through automated causal reasoning.
! DYOR (Do Your Own Research)