Experienced software engineer specializing in AI-native applications, GenAI integrations, cloud architecture, and full-stack development. Hands-on CloudOps experience on AWS with Terraform-managed infrastructure, and hardening CI/CD and release pipelines. Quick learner with a self-driven attitude, passionate about problem-solving and exploring emerging technologies.
Experience
Served as Architect and Developer for research and development on the design and implementation of cloud solutions for banking and financial services.
- Researched and designed the core AI architecture for internal banking project using agent workflow
- Developed and implemented cloud solutions for the finance team on Slack bots and Slack workflow automation
- Developed a PoC and proposed solutions for an internal assistant using GenAI and RAG for banking and financial services
One of four main engineers dedicated to Katalon's first AI-Native Project: Scout Vibe Testing App
- Researched and designed the core AI architecture for Scout project using multi-agent orchestration
- Developed and collaborated with teammates to design and analyze tech stacks to meet requirements
- Collaborated and shared knowledge with other teams about Agentic AI/AI Platform, LLM, RAG, and GenAI
- Adopted state-of-the-art stack including Mastra, AGUI, AgentCore, and MCP
- Worked on Katalon's core AI-powered project, which has an operating cost of 100K USD
- Implemented the billing, credit, and payment system using Stripe, metered credit usage, and lifecycle emails
- Implemented prompt caching, inference optimization, and cost monitoring to reduce LLM inference costs by 30% while maintaining performance
- Built an enterprise Slack assistant on top of RAG and MCP, integrated with multiple data sources (Confluence, Katalon Docs, Forum, Drive) for the entire Katalon internal team, achieving 100% adoption and positive feedback
Cloud engineering intern focused on serverless architecture and AI services integration
- Built serverless web applications using API Gateway, Lambda, and Amplify
- Developed RAG-based GenAI chatbot using Bedrock, OpenSearch, and DynamoDB
- Hands-on experience with VPC, EC2, S3, ECS, CloudWatch, and CodePipeline
- Supported customer pricing calculations using AWS Pricing Calculator
- Participated in cloud architecture workshops and best practices training
Teaching web development to students with a focus on practical project implementation
- Frontend: Instructed HTML5, CSS3, TailwindCSS, JavaScript, TypeScript, and ReactJS
- Backend: Taught MySQL, ExpressJS, NestJS, Prisma, Docker, and CI/CD
- Designed and graded practical projects for student assessment
- Mentored students through final capstone projects
- Achieved 95% student satisfaction rating
Projects
Live at scoutqa.ai — a 'vibe-testing' assistant specialized for automating and testing web pages created by 'vibe coding,' especially for Lovable and Replit. The product's intent is to allow individuals with little to no coding experience to ship their prototype products to production with less friction and fewer unpredictable incidents.
- Successfully deployed the multi-agent orchestration architecture in production
- Integrated latest AWS Services AgentCore and Nova Act for AI-powered features
- Developed the core features by customizing and using a state-of-the-art stack: AG-UI protocol, LiveView browser, and DCV SDK
- Integrated the computer vision tool from Anthropic Computer Use while it was still in preview; it is now official
- AI-Native App built with the Mastra agent framework, CopilotKit, and NextJS
- R&D role covering both research and development of the project, with an operating cost of 100K USD
- Designed and shipped a credit-based usage mechanism for metering LLM consumption and surfacing cost back to end users, using feature flags to gate rollout across user segments
- Integrated Stripe for subscription billing and metered credit purchases, wiring lifecycle emails (welcome, weekly reminders, unsubscribe) and customer-facing billing flows
- Ran A/B experiments on new product features with PostHog, analyzing user behavior, conversion funnels, and traffic segmentation to inform product decisions
Built multiple N8N workflows for automating and auditing security compliance in IT following two security standards: PCI DSS and ISO 27001:2022. Teamed up with 4 others. I contributed to building an MCP Server for crawling and searching information on the official AWS website. My team and I advanced to the final round at VPBank Hackathon 2025 with this project.
- Automated auditing to check compliance with the latest or any updated security standard
- Integrated N8N, GenAI, and MCP for AI-powered features
- Achieved 99.9% time savings compared with manual traditional auditing
- I was also a presenter at the Final Round of the hackathon that took place in VPBank Hanoi
Multi-source integration for ingestion and handling; an AI Slack chatbot built on top of the MCP stack
- Deployed with zero downtime using Lambda functions and rolling updates
- Integrated MCP and wrote Terraform for AI-powered features
- Achieved 100% adoption by internal users
- Implemented in the HonoJS framework and deployed on Lambda, integrated with Slack API
- Ingestion from multiple sources: Confluence, Jira, Drive, Forums, Katalon Docs
Education
Certificates
Awards
Developing an AI-powered solution for automating and designing AWS Deployment Architectures in banking
Selected from 1,200+ candidates to compete in the Security Challenge, developing an Auto Audit Framework using GenAI and n8n
Winner of the technology innovation competition co-hosted by HUTECH and HCA
Champion of the English proficiency competition for tech students
Recognized for innovative banking solution development
Selected among top startup teams for innovative business solution
Skills
References
Principal Software Engineer at Katalon - Email: son.dao@katalon.com