Itikela Bhaskar
About

How I work.

I'm an undergraduate at IIIT Hyderabad pursuing a combined B.Tech and M.S. by Research in Computer Science and Engineering. Most of my recent focus is in the Video-Language Group at CVIT: multimodal models, video understanding, and turning research ideas into reproducible experiments.

I also ship backend code in a startup context (APIs, integrations, cloud), and I use hackathons as a forcing function to go deep fast — on-device inference, geospatial ML under tiny labels, and agentic pipelines on GCP.

This site is where I publish longer case studies when I have a story worth the length; shorter notes live under "Notes."

Highlights
  • Hackathons: seven wins including on-device AI (Qualcomm Megathon), agentic data quality (Lloyds × GCP), and flood mapping from SAR (AISE Hack).
  • Projects: Med Veda (on-device MedGemma Android), cross-lingual QA pipeline, xv6 demand paging.
  • Certifications: OpenCV with Python; Google Cybersecurity; Google AI Essentials.

Timeline

Apr 2025 — Present

Undergraduate Researcher, Video-Language Group

CVIT, IIIT Hyderabad

Multimodal learning, video understanding, and vision-language models with Prof. Makarand Tapaswi.

2025

Backend Developer

Hustlr (startup)

API design, server-side logic, and cloud integration for an early-stage product.

Jul 2023 — May 2028

B.Tech + M.S. by Research in CSE

IIIT Hyderabad

UGEE rank 77. Coursework and research across ML, systems, and vision.

2022

Higher secondary

SAHITTII Junior College

95.4%

2020

Secondary

Viswabharati High School

99.16%

Stack

Languages
  • C
  • C++
  • Python
  • JavaScript
  • SQL
AI / ML & vision
  • PyTorch
  • TensorFlow
  • scikit-learn
  • OpenCV
  • MediaPipe
Systems & infra
  • Linux
  • POSIX
  • TCP/UDP
  • Sockets
  • Docker
  • Git
  • Redis
Web & data
  • React
  • Next.js
  • Node.js
  • React Native
  • Tailwind
  • PostgreSQL
  • Supabase
  • GCP
  • BigQuery