Hey, I'm Shiyan
I'm a senior software engineer based in San Francisco. I'm currently at EliseAI building Apollo, an AI agent for property management. Previously, I ran Sapient Studio, a design and engineering studio that partnered with companies to create impactful digital experiences. I studied computer engineering and business at Queen's University, and I'm a Cansbridge Fellow.
Work
Early engineer at a YC-backed AI-native loan origination startup, working onsite with top U.S. banks to architect and ship AI-powered lending workflows. Built an ML pipeline for automated bank statement analysis to detect spending patterns and improve underwriting decisions, and owned key third-party integrations end-to-end.
Founded and grew a design & engineering studio that partnered with companies to create impactful digital experiences. Built full-stack applications that helped generate $1M+ in ARR. Expanded the team to three engineers and designers, worked with a wide variety of clients, secured funding, and completed two accelerator programs.
Led the development of core infrastructure, a reusable codebase, and shared utilities for 20+ enterprise LLM RAG and agentic AI use cases. Built GenAI use cases and an alerting platform to enhance LLM observability and system responsiveness.
Shipped features that helped users trade and invest in crypto, including auto-buy and staking features. Implemented automated mobile testing frameworks and built alerting systems to prevent unexpected API costs.
Optimized BFCM performance and resiliency by improving queries, refactoring, and implementing load tests. Developed AccessScope and Metafield prototypes to enforce data access limits and improve developer experience.
Increased data collection at an RBC Ventures carbon footprint startup by building OCR receipt scanning, grocery list, and onboarding features. Conducted user interviews and A/B tests to refine features and boost engagement.
Improved energy cost efficiency by broadcasting surges to 1M+ thermostats using Pub/Sub. Mitigated data exposure risks by optimizing queries on the public-facing Demand Response API.
Automated complex tax document translations with GPT-3. Built a sentiment analysis web app to identify user trends from 1K+ government benefit reviews. Explored various models as part of the AI research team.
Automated Excel analysis for 3K+ real estate properties by writing Python scripts. Researched and implemented a 3D imaging and VR facility scanning pilot with a team of six engineers.