Agentic AI / Quant
Vibe Quant
Autonomous agentic alpha discovery framework — LLMs orchestrate research, Qdrant stores vector memory, and a Rust-backed engine validates backtests.
LLM AgentsQdrantRustBacktesting
AI agents · finance systems · data engineering
Hi, I'm Sim Sze Yu.
I make ideas
executable.
AI-native builder working across data engineering, finance systems, and agentic workflows.
Sze Yu OSidentity
4 core areas
AI-native builder
FocusAI agents, finance systems, data engineering
BaseComputer Science, Universiti Malaya
ModeWorkflow-first product builder
practical software
What I build
AgentsRAG, MCP tools, orchestration
FinanceMarket data, backtesting, decision support
DataETL pipelines, quality, cloud workflows
18 hackathons
Proof of work
UM DatathonChampion, Bitcoin quant strategy
Alibaba Cloud2nd Runner Up, FundSight AI
KitaHackFinance planner and Medimate
Medium notes
Writing loop
Pydantic AIAgent development guide
AI workflowHow to get more out from AI
CI/CDuv cache and GitHub Actions notes
01 / Featured Work
Work shown on equal ground.
Each project is treated as a proof point, not a hierarchy. The goal is to make the system shape, technical direction, and links easy to scan.
02 / Capabilities
Skills grouped by what they enable.
Not a keyword cloud. Each group answers one question: what kind of work can you do with this?
Agentic AI
Building autonomous agents that reason, use tools, maintain memory, and produce validated structured outputs.
LLM AgentsMCPPydanticAIRAGQdrantMemoryGuardrailsEvaluation
Data Engineering
Production pipelines, warehouse architecture, and data quality systems built to survive 3am incidents.
AirflowSnowflakedbtKubernetesOSS / S3CI/CDData Quality
Quant & Finance
Systematic research, signal design, backtesting infrastructure, and market reasoning systems.
BacktestingMarket DataStock MetricsSignalsExpert SystemsPython
Product & Web
Full-stack delivery from FastAPI backends to Next.js frontends — whatever it takes to ship.
Next.jsReactTypeScriptTailwindFastAPIshadcn/ui