Jinghao (Adam) Chen

I came to data from the business side, with extensive experience across financial advisory, immigration consulting and e-commerce. That work taught me what a stakeholder actually needs from a report. Now I leverage the technical skills I have built since to support business decisions backed by data, from ingestion and medallion-architecture warehouses to tested transformations and semantic models that hold up when someone checks the numbers.

The clearest highlight of that transition came as a research assistant at Langara College, where I built the data layer behind a community-services app for shelter staff. That work contributed to a project later awarded a $359,928 NSERC grant.

The financial-market analytics platform is a different project, and one I started on my own after the Langara program to teach myself data engineering. I built it first in dbt, Airflow and Streamlit on Google Cloud. After earning the Fabric Analytics Engineer certification, I ported the whole pipeline to Microsoft Fabric and rebuilt the reporting in Power BI. I then added a Tableau Public dashboard fed by Google Sheets through Apps Script. Both versions are still running.

Toolkit

Languages

SQL Python Pandas R DAX

BI

Microsoft Power BI Tableau Streamlit Excel / Power Query

Transform

Microsoft Fabric dbt Apache Airflow 3 dlt

Cloud & ML

Google Cloud Docker GitHub Actions scikit-learn LangChain NLTK

Financial market analytics platform

Microsoft Fabric · Power BI · Pyspark · Google Cloud · Airflow · dbt · Streamlit

An end-to-end ELT platform ingesting daily prices, company profiles and news sentiment for the three most actively traded stocks. Built twice: a Bronze–Silver–Gold medallion architecture of four PySpark notebooks in Microsoft Fabric, and the original open-source stack on GCP.

• Developed a financial market analytics platform to analyze the day’s three most actively traded stocks, identify potentialdrivers of sharp price and volume movements, and present market data through Power BI.
• Extracted and transformed raw API data into analysis-ready Delta tables using PySpark and a medallion architecture in Microsoft Fabric, applying data cleaning, validation, and aggregation.
• Designed a Power BI semantic model and DAX measures comparing closing prices and trading volumes with 100-day averages to highlight unusual activity and support analysis of market trends.
• Built and deployed a GCP version of the platform, running Airflow and dbt in Docker on a virtual machine for daily incremental PostgreSQL updates and hosting a Streamlit dashboard on Cloud Run.

• Live Streamlit app ↗ (opens in a new tab)
• Airflow/dbt backend repo↗ (opens in a new tab)
• Fabric repo ↗ (opens in a new tab)
• Power BI Dashboard ↗ (opens in a new tab)

Real-Time Access to Essential Services

Streamlit · Python · SQL · NLTK · LangChain · Heroku · Github Action

A shelter-finding web app for community services, built as a funded research project at Langara College's Applied Research Centre. The project was awarded $359,928 from NSERC's College and Community Social Innovation Fund.

• Independently developed and deployed a Streamlit application for an external client, enabling community members to locate nearby shelters and view real-time bed availability.
• Used NLTK to clean and tokenize shelter reviews, extract recurring keywords, and analyze rating distributions in shelter-specific manager dashboards, restricting each manager to data for their assigned shelter.
• Built an LLM-to-SQL chatbot that converted natural-language questions into database queries and returned user-friendly answers, with guardrails that rejected out-of-scope requests.
• Translated client feedback into an interactive geolocation map with color-coded availability indicators and database-backed, role-based workflows for managers and administrators to update shelter capacity, community events, and user access.

• Demo video ↗ (opens in a new tab)
• NSERC award record ↗ (opens in a new tab)
• Mentioned in Langara news ↗ (opens in a new tab)

AI-powered HVAC optimisation

Python · sklearn · Tableau · beautifulsoup

Capstone for Langara College's facilities department and Prism Engineering: detecting anomalies in HVAC sensor data and forecasting temperature 15 minutes ahead to support energy-efficient control.

• Analyzed real-time HVAC sensor data from five air-handling units serving 117 zones in Langara College’s Building B, using Python to identify operational anomalies and energy-efficiency opportunities.
• Developed a CNN-LSTM model to forecast 15-minute zone-temperature changes and support data-driven HVAC set-point recommendations that balance energy efficiency with occupant comfort.
• Translated analytical and model outputs into anomaly heat maps and an interactive Tableau dashboard, presenting decision-ready findings to Langara Facilities and Prism Engineering in support of operational improvements.
• Won the Innovative Award at Langara's Applied Research Day 2025.

• Project site ↗ (opens in a new tab)
• Project post on LinkedIn ↗ (opens in a new tab)

1. Research Assistant

   Jan 2025 – Jul 2025

   Langara College Applied Research Centre · Vancouver

   • Analyzed stakeholder requirements, researched feasible solutions, and presented a phased development plan with realistic timelines to extend a coursework prototype into a community-services application, contributing to a research project later awarded a $359,928 NSERC grant.
   • Applied Agile principles to iteratively redesign the dashboard after stakeholder discussions revealed heavy workloads and staffing shortages, enabling shelter staff to complete bed availability updates in about one minute.
   • Developed an AI-powered chatbot that enabled users to ask questions and receive shelter information in plain language, with safeguards to reject out-of-scope requests.
   • Conducted a cost analysis of database and cloud hosting options against application requirements, leveraging GitHub Student Developer Pack benefits to keep the application at little to no cost throughout the research period.
   • Wrote technical documentation in GitHub to support project handover and help future students understand the application and continue its development.

2. Course Assistant

   Sep 2024 – Jul 2025

   Langara College · Vancouver

   • Marked two computer science units, giving detailed written feedback on assignments.
   • Planned and hosted the capstone competition event, coordinating student presentations and industry judges.

3. Travel Counselor

   May 2024 – Aug 2024

   Discover Surrey · Surrey, BC

   • Collected and reported event attendance and visitor data used for tourism planning and small-business support.

Earlier career: finance, trading and operations (2016–2022)

• Retail Stock Trader · 2020 – 2022 Backtested swing, day and options strategies; technical analysis for risk-managed position sizing.
• Immigration Consultant · CHENG QI (Singapore) Pte. Ltd. · 2019 – 2021 End-to-end case management through changing regulatory requirements.
• Financial Consultant · IPP Financial Advisers · 2018 – 2020 Portfolio reviews and data-driven risk assessments; built WeChat acquisition campaigns.
• Ecommerce Manager · Hao Jun Private Ltd · 2017 – 2018 Omnichannel sales analysis for product assortment; supplier negotiation with China-based partners.
• Business Development Executive · Guangzhou Rui He Fu Investment Management · 2016 Roadshow project targeting IPO candidates; provider negotiations and cost savings.

1. Post-Degree Diploma, Data Analytics

   Vancouver · 2024 – 2025

   Langara College, graduated with Distinction

2. Bachelor of International and Global Studies

   Sydney · 2012 – 2015

   University of Sydney · Majors: International Business & Political Economy

Certifications

Grouped by the kind of work they map to, rather than the order I earned them.

Publication & awards

• Innovative Award, Applied Research Day 2025, Langara College For the AI-powered HVAC energy optimisation capstone. See the Applied Research Day post ↗ (opens in a new tab)
• Second place, Langara Open Student Scholar Prize 2024 For “Employing Random Sampling Techniques on Machine Learning Models”. See the 2024 winners list ↗ (opens in a new tab)

Email is the fastest way to reach me: no queue, no backfill, guaranteed delivery.