Abhinav Siddharth Tanniru
I build at the intersection of hardware constraints and intelligent systems — from UAV firmware and image-quality pipelines at Qualcomm, to published research on RL-driven edge computing. Most interesting to me: the problems that only emerge when software has to run on real, constrained hardware.
Professional Experience
R&D Software Engineer
Dec 2025 - Present- Led camera triggering, synchronization, data integrity, and calibration workflows across diverse UAV camera payloads.
- Improved mapping accuracy and 3D reconstruction reliability through targeted software and firmware enhancements.
Software Engineer
Aug 2025 - Nov 2025- Developed scalable MERN applications with improved frontend responsiveness and backend API throughput.
Systems Engineer Intern
Feb 2025 - Aug 2025- Conducted objective (PSNR/SSIM) and subjective image quality evaluations across varied lighting and noise conditions.
- Benchmarked multiple codecs for image transmission efficiency and system tradeoffs.
- Collaborated with hardware teams to deliver 10-15% image-quality gains while reducing AR/VR power requirements.
Data Science Intern
Feb 2024 - Jul 2024- Built a GST returns data pipeline (retrieval, preprocessing, and storage), improving documentation and execution efficiency.
- Optimized GST data collection flows, contributing to at least 30% cost reduction.
- Led migration of an API from monolithic architecture to service-oriented design.
Research Assistant
Jan 2022 - July 2025- Conducted vehicular edge computing research spanning RL and deep learning for embedded and IoT systems.
- Teaching assistant for Applied Optimization, Real Analysis, and Linear Algebra.
Research & Publications
Projects
Codec Explorer
Built an interactive codec lab that makes transform-based compression internals observable in real time.
Lemma
Built a structured math study tracker with PDF viewing, topic notes, problem logging, and Supabase-backed auth — all in a single cohesive app.
Y86 Lab
Built an interactive browser-based visualizer for a 5-stage pipelined Y86-64 CPU, making pipeline state observable step by step.
Custom OS Shell
Implemented process, signal, and command parsing internals to emulate core Unix shell behavior.
Data2Vis Transformer
Built a sequence model that generated valid Vega-Lite specs directly from raw tabular inputs.
E-Learning Platform
Delivered an LMS with role-based access, media delivery, and dashboard-level observability.
Seq2Seq Translator
Built a GRU-based encoder-decoder translation model trained on 7,000 sentence pairs, achieving 50% test accuracy.
Housing Prices ML
Built end-to-end regression pipelines for California housing-price prediction with feature engineering and cross-validation.
Education
International Institute of Information Technology, Hyderabad
Graduated July 2025B.Tech (Honors) + Master of Science, Electronics
CGPA: 9.0
Let's connect
I'm selectively open to interesting problems in embedded systems, edge AI, and computer vision research. Whether it's a collaboration, a role, or just an interesting idea — I'm happy to talk.