Syed Mohammad Qasim
I am a PhD candidate in Computer Engineering at Boston University, advised by Prof. Ayşe K. Coskun and co-advised by Prof. Gianluca Stringhini. I work at the intersection of distributed systems and machine learning: making large-scale cloud platforms easier to observe, debug, and keep reliable, using ML-based anomaly detection and LLM agents that reason over distributed traces.
Most recently, I built DSTBench (IC2E 2026), a benchmark for evaluating LLM agents on trace-driven performance debugging in microservices, in collaboration with researchers at Google. Earlier work includes online probabilistic tracing (Best Paper Award, IC2E 2024) and a production ML anomaly detection deployment on Chameleon Cloud's OpenStack services (IC2E 2025). This summer I interned at Uber, improving the precision of alert-triggered automated rollbacks for deployments.
Before my PhD, I spent five years in industry — leading platform engineering at State Street and building enterprise microservices and blockchain systems at Thoughtworks.
News
- Aug 2026Wrapped up my PhD software engineering internship at Uber, working on deployment safety and automated rollbacks.
- 2026DSTBench accepted to IC2E 2026.
- Sep 2025Presented Lessons Learned from Anomaly Detection in Chameleon Cloud at IC2E 2025.
- Sep 2024🏆 Unleashing Performance Insights with Online Probabilistic Tracing won the Best Paper Award at IC2E 2024.
Publications
Experience
- Improved the precision of alert-triggered automated rollbacks, reducing unnecessary rollbacks caused by low-confidence alerts.
- Enriched rollback decisions with the service dependency graph and surrounding metric context to attribute alerts to the deployment that actually caused them.
- Build ML- and LLM-based tools that improve observability and performance debugging in large-scale cloud platforms.
- Developed and deployed an ML-based anomaly detection system for Chameleon Cloud's OpenStack services, improving observability and fault detection.
- Automated ML model development and deployment (DataRobot) to fix data drift and model decay; automated upgrades across 100+ servers with Ansible Tower.
- Built a front-to-back client servicing platform on API Gateway: 2-way TLS, IP whitelisting, OAuth2, and API-key auth, plus automated maintenance including a virus-scanning microservice and cloud backups.
- Built a decentralized oil-trading consortium app on Ethereum-Quorum with Solidity smart contracts, deployed on AWS ECS.
- Built scalable Spring Boot microservices for real-time Kafka ingestion, including a custom Hibernate object-mapper for high-throughput nested payloads.
Education
Skills
Languages: Python, Java, C++, SQL, JavaScript
Systems & infrastructure: Kubernetes/OpenShift, OpenStack, Kafka, Elasticsearch, Spring, Ansible
ML & LLMs: scikit-learn, TensorFlow, LLM agents & tooling (Ollama), distributed tracing & observability