Beijing, China | Remote OK
Email: phnasis@126.com
I am an AI & Data Systems Engineer with 14+ years of experience building large-scale, high-availability data platforms in financial and regulated environments.
My work focuses on turning complex systems into intelligent, reliable, and scalable products — especially at the intersection of:
- AI Agents & RAG pipelines
- Real-time data infrastructure
- Workflow orchestration & optimization
- Web3-native data & oracle systems
I enjoy working on problems where data, system design, and intelligence meet real production constraints.
- AI-assisted scheduling & decision systems
- Log intelligence & failure explanation using LLMs
- RAG pipelines for structured + unstructured data
- Feature engineering on DAGs, execution traces, and system metrics
- Reinforcement-learning–inspired optimization (DDQN / IDDQN)
- High-throughput ETL & streaming pipelines
- Kafka / RabbitMQ
- PostgreSQL / MySQL / Oracle
- Elasticsearch (large-scale indexing & search)
- Low-latency, high-availability system design
- Workflow orchestration (DolphinScheduler)
- Multi-cluster & multi-tenant architectures
- Observability, performance profiling, fault diagnosis
- Financial-grade stability & reliability engineering
Java · Python · Kafka · Redis · Elasticsearch · Docker · Linux
Group-level scheduling platform serving regulatory reporting, AML, actuarial, risk control, and data platform teams.
What I built
- Multi-cluster scheduling architecture (commercial + open-source clusters)
- AI-assisted scheduling to replace rule-based execution
- Execution log intelligence for failure analysis and root-cause explanation
- DAG semantic modeling (dependencies, resources, duration)
Impact
- 40–50% reduction in critical workflow MAKESPAN
- Significant improvement in cluster resource utilization
- 1000+ internal users, million-level task executions
- Zero major production incidents
Unified data access layer for 36 provinces, serving over 1M internal users.
Highlights
-
10M daily queries
- 99.92% requests completed under 1s
- Zero major incidents since launch
- Cross-team architecture leadership (60+ engineers)
I am particularly interested in:
- Agentic Oracles & Social Signal Processing
- Prediction Markets & Market Intelligence
- On-chain / Off-chain data pipelines
- Solana ecosystem infrastructure
If you're building at the intersection of AI, data infrastructure, and Web3,
I’d love to chat.
📧 phnasis@126.com
🌍 Open to remote & early-stage teams