M. Tanzil Idrisi

Software Engineer Intern · San Francisco, California

• Built a scalable ads auction system in Go handling 10M+ daily requests at p95 under 120ms.
• Maintained 99.9% auction availability across multi-AZ Kubernetes deployment and failover.
• Shipped a feature-flag service powering checkout experiments with automated rollout controls.
• Reduced PostgreSQL load 40% with Redis caching for sub-millisecond metadata lookups.
• Processed billions of events with Airflow pipelines, distributed tracing, and automated recovery.

Software Engineer Fellow · Remote

• Selected from 400+ applicants for a competitive 10-week program built around hands-on projects, interview preparation, and technical workshops.
• Received one-on-one mentorship from Google engineers and collaborated with peers on production practices.

Undergraduate Research Fellow · Lübeck, Germany

• Innovated a noise-aware 3D U-Net, improving PET image quality by 15% and surpassing baseline U-Net performance.
• Worked on AG-PET, an anatomically guided multimodal U-Net for PET denoising.
• Constructed a patch-based NAFNET noise-aware fusion network.
• Built a self-supervised diffusion model made noise-aware to improve denoising accuracy.
• Presented findings at the IEEE NSS Medical Imaging Conference 2024 in Tampa, Florida.

supervisors · Ezzat Elmoujarkach, Magdalena Rafecas

Machine Learning Research Intern, School of Computer Science · Pittsburgh, Pennsylvania

• Developed a Python mobile system for 3D face reconstruction from 2D images, supporting cancer and cosmetic surgery planning.
• Leveraged a GAN trained with PyTorch, achieving a 90.2% accuracy rate on the LFW dataset.
• Used DLib to identify 68 key facial landmarks, with SciPy, PyMesh, and Matplotlib.

supervisor · Dr. Ganesh Mani

Data Science Engineer · Beloit, Wisconsin

• Developed a Python desktop application for task automation, saving $10K+ by producing custom financial aid letters in house rather than through an outsourced service.

Software Engineer Intern · Menlo Park, California

• Built Vision Transformer optimizations for large-scale video understanding across Meta production systems.
• Designed adaptive token sampling in PyTorch and integrated it into Reels video understanding pipelines.
• Reduced transformer inference latency and FLOPs by 25%, cutting production compute costs by 10%.

Co-Founder & CEO · United States, remote

• Provided custom AI solutions to companies, from marketing and branding to data-driven tooling.
• Focused primarily on the financial and e-commerce sectors.
• Specialized in helping small and medium enterprises grow efficiently and cost-effectively.