Current work
Applied AI web systems
At Microsoft AI, I build enhanced page features, web document features for grounding models, and URL normalization systems that improve resource efficiency, deduplication, and product quality.
PhD in Computer Science · Security, Privacy, IoT, and Systems
Security and privacy research, built for real systems.
I’m Dilawer Ahmed, a software engineer at Microsoft AI and a researcher focused on security, privacy, and production systems. My work spans voice assistants, IoT measurement, web privacy, and the infrastructure required to turn research ideas into deployable systems.
Focus
What I work on
A quick view of the engineering, research, and academic threads that shape my work.
Research
Security, privacy, and the web
My research focuses on security, privacy, voice assistants, IoT, and web platforms, with an emphasis on making measurement and attack insights meaningful for real systems.
Background
Engineering with a research lens
I completed a PhD in Computer Science at NC State and earned a BS in Computer Science from LUMS, which shapes how I balance rigor, usability, and production constraints.
Career
Experience
Research, product, and engineering work across industry and academia.
Software Engineer II
At Microsoft AI, I build enhanced page features and web document features that improve how models ground on the open web. I also work on URL normalization and related platform capabilities that reduce duplication, improve resource efficiency, and strengthen product quality.
Software Engineer Intern
At Microsoft, I worked on location intelligence systems for advertising, redesigning an offline pipeline in Azure Data Lake to improve recall by up to 50% and increase downstream impact. I also shipped a C++ online serving path and partnered with the Bing Maps team on reverse geocoding, reducing compute cost by 87% while increasing coverage by 25%.
Software Engineer Intern
At Google Cloud, I explored GPU virtualization strategies for machine learning acceleration and built a C++ performance projection tool to estimate remote GPU latency and bandwidth with lower simulator overhead. I also helped drive a protocol buffer dependency migration that reduced maintenance friction and improved internal developer workflows.
At KalPay, I led engineering for a BNPL platform, defining the product architecture and guiding execution across the merchant, consumer, and admin applications. The role combined early-stage product thinking with hands-on delivery across the company’s core fintech stack.
Software Engineer
At Technologies of People Initiative Lab, I built research prototypes and data systems for applied computing projects, translating exploratory ideas into usable systems. My work spanned full-stack implementation, experimentation, and technical support for the lab’s broader research agenda.
Projects
Selected Projects
A mix of research programs and production systems work across voice assistants, IoT, web privacy, ads, infrastructure, and applied ML.
Voice Assistant Security & Privacy
Measuring, attacking, and defending modern voice assistant ecosystems, from traffic fingerprinting to scalable policy analysis and deployable defenses.
Location Intelligence Systems
Production engineering for large-scale location processing, delivery, and ranking systems, with a focus on reliability, quality, and computational efficiency.
Privacy-preserving Vision Systems
Designing transformations that preserve utility in camera and video data while making re-identification and leakage substantially harder.
Cloud GPU Systems
Tooling and systems work for reasoning about remote GPU performance, virtualization, and infrastructure behavior in ML-oriented cloud environments.
Web Advertising & Privacy Controls
Studying how advertising ecosystems, exception lists, and privacy tools interact in practice, with a focus on the real privacy cost of defaults.
IoT Privacy & Device Fingerprinting
Understanding how encrypted traffic reveals device identity and usage patterns, and designing measurements that better capture open-world privacy risk in IoT ecosystems.
Research
Latest Publications
Recent conference papers, posters, and research outputs.
Spying through your voice assistants: Realistic voice command fingerprinting
We show that multiple voice assistant platforms can be fingerprinting equally effectively. We also show that the fingerprinting process can be performed remotely mixed with traffic from other devices. Adding additional features such as flow and burst based features can also increase fingerprinting performance
Analyzing the Feasibility and Generalizability of Fingerprinting Internet of Things Devices
We show that not only is it possible to effectively fingerprint 188 IoT devices (with over 97% accuracy), but also to do so even with multiple instances of the same make-and-model device. We also analyze the extent to which temporal, spatial and data-collection- methodology differences impact fingerprinting accuracy. Our analysis sheds light on features that are more ro- bust against varying conditions
Connect
Contact
Based in Mountain View and open to conversations about security, privacy, applied research, and engineering leadership.
dilawer11@gmail.com
Phone
+1 (nine 84) three 89 376 seven
Location
Mountain View, CA
United States