Chandana Kiran

AI Researcher

Austin, Texas, United States4 yrs 3 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Optimized AI inference for generative models at AMD.
Developed a full-stack solution reducing task time by 95%.
Led a capstone project for neurodiverse STEM learners.

Stackforce AI infers this person is a skilled AI/ML Engineer with strong software development capabilities.

Contact

chandana0699@gmail.com LinkedIn

Skills

Core Skills

Machine LearningAi InferenceSoftware DevelopmentDebuggingFull Stack Development

Other Skills

BootstrapC (Programming Language)C#C++Cascading Style Sheets (CSS)Cloud ComputingCryptographyCybersecurityData PipelinesData ScienceDeep LearningDjangoEmbedded SoftwareFront-End DevelopmentHTML

About

I bring a deep curiosity and a strong focus on execution. I take pride in going beyond the ask and delivering with ownership, while building solutions that are efficient and built to scale. With 2 to 3 years of industry experience, I’ve worked on AI inference optimization, generative models, and system-level engineering. At AMD I optimized inference for generative AI models, focusing on speed, scalability and deployment. At Qualcomm, I tuned modem-level software to enhance performance across embedded platforms. As a graduate student in Computer Science at Virginia Tech, I led a capstone project using RAG to personalize STEM content for neurodiverse learners, combining technical innovation with social impact. I’m passionate about Machine learning, Deep learning, and Generative AI, especially building AI systems that are not only powerful but also practical and purposeful. Open to meaningful conversations and collaboration, feel free to reach out!

Experience

Amd

2 roles

AI Inference/ Performance Engineering

Jul 2025 – Present · 8 mos · Austin, Texas Metropolitan Area · Hybrid

AI Engineer Intern

Aug 2024 – Apr 2025 · 8 mos · Austin, Texas, United States

Built ML and DL models to predict application performance across chip architectures, enabling cost-efficient design decisions.
Analysed chipset data through comprehensive EDA and feature engineering to extract insights and enhance interpretability.
Optimized CPU recommendations for customer workloads, reducing validation time and improving deployment efficiency.
Enhanced AI Inferencing performance for Large Language Models (LLMs) on CPUs through quantization and system configuration optimizations, achieving a significant 11.4x boost in throughput and latency.
Orchestrated deployment of containerized generative AI models (Llama 3) on Kubernetes using TGI server, enhancing scalability.
Designed a deployment pipeline for object detection models reducing the runtime from 28 mins to 3 mins through an optimized multiprocessing approach.

Machine LearningDeep LearningAI InferencePerformance EngineeringKubernetesPython

Virginia tech

Research Assistant/ Instructor - Qualcomm Thinkabit Lab - Virginia Tech

Nov 2023 – Jun 2024 · 7 mos · Virginia, United States · On-site

Mentored students across Virginia in immersive Micro:bit and Arduino-based STEM activities while emphasizing collaborative problem-solving and project based learning.
Led comprehensive analysis of raw data from the Invent Virginia competition, Virginia School extracting valuable insights from large-scale datasets to inform strategic decisions and enhance participant experience.

Qualcomm

3 roles

Software Engineer 2

Promoted

Aug 2022 – Jul 2023 · 11 mos

Worked on modem system software development tasks such as pre-silicon validation, SoC bring up and feature development.
Developed a Python parser and solely contributed to ~105kb of the team’s overall memory reduction target of 250kb.
Worked on debugging multiple modem software stability issues in wearable devices.

Software DevelopmentPythonDebuggingSystem Software Development

Associate Software Engineer 1

Jun 2021 – Jul 2022 · 1 yr 1 mo

Software Engineering Intern

Feb 2021 – Jun 2021 · 4 mos

Designed an interactive website using Full Stack development (Django, Python, Js, MySQL) to automate a laborious 5-day inter-team task, effectively eliminating manual efforts and streamlining operations.
The website seamlessly extracts modem test scenario statistics from RAM dumps, achieving an impressive 95% reduction in task completion time and markedly enhancing overall efficiency.

Full Stack DevelopmentDjangoPythonJavaScriptMySQL

Unisys

Summer Intern

Jun 2019 – Jul 2019 · 1 mo · Bengaluru, Karnataka, India

• Interned at UNISYS for a duration of 2 months while working on the Service Now platform to build a Virtual Agent (IRIS-Chatbot) for a project regarding Knowledge Management.

Tequed labs

Intern

Jul 2018 – Jul 2018 · 0 mo

• Attended a hands-on workshop regarding the Internet of Things and successfully completed projects based on Pothole Detection within a period of 4 weeks.