Welcome to my GitHub. I'm a tech-focused product manager based in Singapore, blending 15+ years of finance and startup experience with hands-on engineering to build tools that make work easier and ideas real.
I’m a very technical product manager and builder with a strong foundation in software architecture, data engineering, and workflow automation. I’ve worked at the intersection of business and technology — leading full-cycle product initiatives, building internal tools, and delivering scalable systems across finance and startup environments.
I bring:
- A hands-on approach to system design and solution architecture
- Deep experience in AI/ML tooling, data platforms, and cloud-native development
- A strong focus on performance, productivity, and real-world maintainability
I’m unapologetically pragmatic when it comes to building things: I use GitHub Copilot to ship working solutions in languages I’m less familiar with, and experiment regularly with tools like Replit AI and other coding assistants to accelerate development. I’m excited about how AI augments human creativity — and I embrace it fully in my workflow.
Whether I’m coding, designing workflows, or leading cross-functional teams, I approach problems with clarity, curiosity, and a bias for clean execution.
🤖 AI & GenAI: OpenAI · RAG · HuggingFace · Pinecone · stability.ai · Github Copilot · Replit
🧱 Languages & Frameworks: Python · FastAPI · TypeScript · Node.js · Qwik · React · Firebase · Java · Kotlin · Spring · Redis · Akka · HazelCast · Postgres · MongoDB · AppSheet · AppScript
📊 Data & Analytics: scikit-learn · SQL · Google Cloud (TensorFlow, BigQuery, Data Studio, Looker, DataFlow) · Apache Beam / Spark · Excel · Access · Tableau · Business Objects · Matlab · R · VBA
☁️ Infra, CI/CD, Cloud: Cloudflare · GCP · Firebase · AWS · Azure · MongoDB Atlas Docker · Kubernetes · Helm · GitHub Actions · GitLab CI · Jenkins · Datadog · Prefect · Airflow
💼 Project & Agile Delivery: Agile · Scrum · Kanban · JIRA · WSJF · Prince2 · MS Project
Other: VS Code · Colab · Databricks · Obsidian · Postman · Linux · Figma
🎥 YouTube: Ctrl Alt Kopi
Tech, Productivity & Kopi ☕
Ctrl Alt Kopi is where I explore technology, automation, and philosophy — all over a good cup of kopi. The channel is built for tech thinkers and tinkerers curious about how tech shapes our work and lives.
What you'll find:
- ✅ Tech & Productivity – AI tools, workflows, automation, and system design
- ✅ Smart Tech & Gadgets – Smart homes, emerging tools, and digital wellness
- ✅ Industry & Philosophy – Product, open-source, and how tech influences culture
- ✅ Personal Journey – Lessons from startups, engineering, and career pivots
🔗 Click to watch: “Replit Agent - AI Coding Assistant”
In this video, I dive into Replit's AI coding assistant — but more importantly, I unpack it like an engineer: evaluating UX, real-world usability, and code maintainability. Even when trying out a no-code/vibe-coding tool, I focus on workflows, limitations, and architectural thinking — the same lens I bring to every product or platform I evaluate.
Topics covered:
- Replit Agent in the wild: messy code, planning gaps, and friction points
- Real-world test of AI-assisted workflows
- Reflections on what’s still missing in AI tools today
Led a team of 15+ engineers (frontend, backend, AI/ML) to build and maintain highly customized webapps for enterprise clients, focused on insights powered by LLMs and internal models. Took ownership of architecture, delivery, and quality across the full stack — stepping in where needed to unblock progress or fill technical gaps.
Key Contributions:
- 🛠 Built a Python/FastAPI microservice from scratch to support natural language querying and consolidated 5+ smaller services into 2 structured and reusable base services.
- 🧱 Architected and future-proofed a highly configurable platform:
- Frontend apps generated via config-based rendering
- Backend highly customised prompting driven by database-controlled logic and flags
- 📊 Developed tooling for prompt engineers to manage and test prompts directly from the database — enabling faster iteration and better prompt governance.
- 🧼 Conducted code reviews and personally refactored code to raise quality and maintainability. Instituted hygiene practices like consistent PR checks and linting.
- 🧰 Stepped in hands-on to fix bugs and deliver features when timelines were at risk or engineers were unavailable.
- 🤖 Ran team workshops on GitHub Copilot to upskill devs and introduce AI-assisted coding practices.
- 🔒 Championed security best practices, led internal compliance reviews, and supported ISO certification preparation.
Over 20+ webapps delivered. Reduced platform clutter by retiring 20+ redundant services and consolidating functionality into stable, well-documented systems.
Drove the development of data products across Southeast Asia, including Onboarding, KYC and risk signals products.
Key Contributions:
- 🧾 Designed detailed API specs using OpenAPI 3.0, providing clarity and enabling quick implementations by engineering.
- 🔧 Built ETL and reverse ETL pipelines to transform and aggregate data in data lake, supporting BI, operations as well as insights generation
- 📊 Modeled data in Looker, building dashboards that informed business reviews, performance tracking, and decision-making across product and sales.
- 🚀 Launched 8+ products across markets (Indonesia, Philippines), supporting geographic and product expansion.
- 🔍 Acted as L2.5 technical support, investigating production issues by reviewing logs in Splunk and metrics in Datadog, triaging root causes for engineering handoff or direct resolution.
- 🛠 Built tools and interfaces for Customer Success, Sales, and Onboarding teams — optimizing workflows, reducing manual work, and improving conversion funnels.
- 🔐 Championed engineering efforts to improve security and scalability — driving containerization, architecture realignment, and implementation of VAPT recommendations.
Led end-to-end product development for M-DAQ’s core Foreign Exchange (FX) engine — from vision and architecture through deployment, monitoring, and continuous improvement. The system was built on an event-driven architecture, processing >50,000 quotes per second with end-to-end latency under 400ms, using Aeron and Akka as the messaging backbone and handling market input via the FIX protocol.
Key Contributions:
- 🔍 Conducted a deep dive of the legacy codebase (primarily Java, with Kotlin components) to reverse-engineer critical business logic. This analysis uncovered redundant pathways and design bottlenecks, leading to a streamlining initiative that reduced latency, lowered operational complexity, and improved reliability.
- 🧯 Acted as L2.5 technical support, using logs and metrics to debug production issues, identify root causes, and flag problematic code paths for fix.
- 🧮 Authored and executed DB migration scripts to correct data inconsistencies and improve schema design for long-term maintainability.
- 🗺 Introduced a formalized Product Vision and Roadmap, aligning engineering efforts with business goals and enabling sharper prioritization.
- 🧪 Delivered a new synthetic product that transformed how FX risk was hedged and priced.
- 🤝 Onboarded a major institutional client (10–15% transaction volume growth) and added 5+ vendor partners to support expansion into 2 new territories.
💡 Note: This profile focuses on my engineering-related work. For full experience, feel free to connect with me.
| Project | Description | Links |
|---|---|---|
| 💰 Kiasu Saver | A Qwik-based app that recommends the best bank interest rates in Singapore based on your savings balances. | GitHub · Live App |
| School Management System | System An internal tool for managing parent/student records, class schedules, and invoicing — built for a tuition centre. | Using AppSheet (private) |
| Spelling Master | A spelling quiz app for children, designed to help them learn through repetition and immediate feedback. | GitHub |
| Fedora HVD | Converts an Azure-hosted Fedora Cloud VM into a hosted virtual desktop (HVD) using Chrome Remote Desktop. | |
| nfs-server | Custom NFS layer that makes GCS buckets usable as low-cost shared storage in a GKE cluster. | GitHub |
| Home Lab | Self-hosted productivity stack (Vaultwarden, Matrix Synapse) on a local machine using Docker Compose. | VaultWarden |
- 📫 Email: [email protected]
- 🧑💻 LinkedIn: linkedin.com/in/tranhailong1986/