Iβm a software engineer focused on building production-minded AI applications using Python, Streamlit, and large language models.
My work emphasizes system design, modularity, and orchestration β not just calling APIs.
Multi-step AI system with explicit orchestration
- Coordinates prompt refinement, content generation, and summarization
- Uses a dedicated orchestration layer to control execution flow
- Modular tool design with shared prompts and centralized LLM configuration
- CI-validated, production-ready structure
π Live App: https://my-bc-st006.streamlit.app
π¦ Repo: https://github.com/Emmatshi/my-bc-st006
Prompt engineering as a first-class tool
- Refines rough user input into clear AI-ready prompts
- Clean generator logic with CI validation
- Demonstrates prompt design impact on downstream generation
π¦ Repo: https://github.com/Emmatshi/my-bc-st005
Focused AI capability with strong UX discipline
- Concise summarization with word counts and copy-to-clipboard
- Defensive secrets handling
- Backported architectural improvements from earlier apps
π¦ Repo: https://github.com/Emmatshi/my-bc-st004
Foundation AI application
- Structured content generation with configurable tone and style
- Clean separation of UI, prompts, and LLM logic
- CI-validated project structure
π¦ Repo: https://github.com/Emmatshi/my-bc-st003
A reusable project template used across all apps:
- Standardized structure
- Poetry-based dependency management
- Secrets handling
- CI setup
π¦ Repo: https://github.com/Emmatshi/streamlit-ai-template
- Designing AI systems, not just demos
- Orchestrating multi-step workflows
- Clean architecture and separation of concerns
- CI, reproducibility, and deployment readiness
- GitHub: https://github.com/Emmatshi
- LinkedIn: (add your LinkedIn URL here)