Location
Corsair - Ho Chi Minh city
Position
Data Engineer
Description
- Collaborated cross-functionally to gather business requirements and deliver AI-driven analytics solutions
- Built and operated end-to-end data pipelines (Docker, Dagster, dbt/Spark, Snowflake/Fabric) on AWS EC2, following Kimball dimensional modeling principles, with Power BI as the reporting layer
- Developed backend services (FastAPI) and MCP servers (FastMCP) using Python for AI-driven applications
- Implemented monitoring and data quality using Grafana, Elementary
- Optimized pipelines, reducing processing time by ~70% (from 6h to <2h) via model refactoring, clustering, and warehouse tuning
- Highlighted Project: Sentiment Analysis
- Built data pipeline for Zendesk, Amazon, Reddit data
- Developed AI chatbot (LangGraph, FastAPI, Postgres)
- Designed LLM workflows for classification, summarization, sentiment scoring
- Highlighted Project: Corsair Marketing Intelligence Solution
- Integrated multi-source data (POS, Keepa, Stackline)
- Engineered features for XGBoost models forecasting market share for Corsair and competitors
- Replaced external solution, saved ~$100K/year
- Highlighted Project: cData Replacement Initiative
- Replaced third-party ingestion with internal pipelines
- Migrated multiple data sources (Shopify, Paypal, Google Analytics) with no business disruption
- Reduced platform cost by ~$20K/year
- Highlighted Project: Corsair CDP MCP Server
- Architected the MCP server connecting Claude to 10+ live data sources via direct API
- Integrated the semantic layer into the MCP server architecture
- Engineered two-layer skill system: platform skills (how to query each source) + analytics skills (what analysis to apply)
- Secured via Azure Entra ID, Key Vault, and Managed Identity
Skills
- Coding language: Python, SQL, JavaScript, TypeScript
- Database: PostgreSQL, Redis, Oracle
- Data Warehouse: Snowflake, Microsoft Fabric, Google BigQuery
- API: REST, GraphQL, Webhook
- ETL Tools: Orchestration (Dagster), Intergration (Airbyte), Transformations (dbt, Spark)
- AI techniques: LLMs, LangGraph, MCP, RAG, Prompt engineering
- Framework: FastAPI, FastMCP, Streamlit
- BI Tools: Power BI
- Monitoring: Grafana, Elementary
- Others: Docker, Linux, Git, Postman, AWS, Jira, Excel
- Soft skills: Agile - Scrum