Varun Sahni — Software Engineer

I am a Software Engineer specializing in Machine Learning Infrastructure, generative AI systems, and scalable backend architectures. With a strong foundation in C, C++, Python, and Java, I build the distributed pipelines that make complex ML models performant, measurable, and production-ready. Currently at Qualcomm, I architect AI infrastructure and evaluation pipelines. I lead the development of "Code Buddy," an enterprise-level, agentic AI workflow that automates 60,000+ code reviews per month. My work centers on designing intelligent AI agents capable of iterative verification. By building dependency graphs to map function hierarchies and utilizing Tree-sitter for AST parsing, these systems can trace the ripple effects of code changes with high precision. Furthermore, I implement custom RAG architectures to analyze complex spec and design documents, ensuring LLMs are grounded in proprietary engineering standards. Previously at AMD, I focused on high-performance computing and distributed data pipelines, optimizing multi-threaded Python scripts to process massive hardware tracing datasets (250GB+) and resolving severe I/O bottlenecks. Core Competencies: • AI/ML Infra: RAG Pipelines, Multi-Agent Systems, Agentic Workflows, LLM Evaluation • Languages & APIs: C, C++, Python, Java, SQL, REST API design • Systems & Cloud: Kubernetes, Docker, Distributed Data Pipelines, Redis, CI/CD • Engineering: High-Volume Data Processing, Multithreading, Backend Architecture Portfolio Website: https://www.varunsahni.in/

Stackforce AI infers this person is a Backend-heavy Fullstack Engineer specializing in AI/ML Infrastructure and Augmented Reality.

Location: Hyderabad, Telangana, India

Experience: 2 yrs 4 mos

Skills

Generative Ai
Ai/ml Infra
Data Pipelines
Large Language Models (llm)
Kubernetes
High Performance Computing (hpc)
Augmented Reality (ar)
Leadership
Mixed Reality
Android Development

Career Highlights

Architected enterprise-level AI workflows at Qualcomm.
Optimized data processing runtime from 7 days to 45 minutes.
Led development of generative AI systems and pipelines.

Work Experience

Qualcomm

Software Engineer (7 mos)

Associate Software Engineer (1 yr 4 mos)

AMD

Intern (5 mos)

Kalipatnapu AR/VR Lab

President (1 yr 1 mo)

Standard Chartered Bank

Summer Intern (2 mos)

Technische Universität Braunschweig

Technical Researcher (4 mos)

Invoxel Technologies

Summer Intern (1 mo)

iWorksLab

Summer Intern (1 mo)

Education

Bachelor of Engineering - BE at Birla Institute of Technology and Science, Pilani

PCM - Computer Science at Manav Sthali School - India

General at Manav Sthali School - India

Varun Sahni

Software Engineer

Hyderabad, Telangana, India2 yrs 4 mos experience

Most Likely To SwitchAI Enabled

Key Highlights

Architected enterprise-level AI workflows at Qualcomm.
Optimized data processing runtime from 7 days to 45 minutes.
Led development of generative AI systems and pipelines.

Stackforce AI infers this person is a Backend-heavy Fullstack Engineer specializing in AI/ML Infrastructure and Augmented Reality.

Contact

Skills

Core Skills

Generative AiAi/ml InfraData PipelinesLarge Language Models (llm)KubernetesHigh Performance Computing (hpc)Augmented Reality (ar)LeadershipMixed RealityAndroid Development

Other Skills

Python (Programming Language)Agentic WorkflowsMLOpsRetrieval-Augmented Generation (RAG)C/C++Dependency GraphsTree-sitterAST ParsingJavaScript4GRancher Kubernetes ManagementC (Programming Language)RedisMongoDBLTE

About

Experience

2 yrs 4 mos

Total Experience

1 yr 2 mos

Average Tenure

1 yr 11 mos

Current Experience

Qualcomm

2 roles

Software Engineer

Promoted

Nov 2025 – Present · 7 mos · On-site

Architected and deployed "Code Buddy," an enterprise-level agentic AI workflow integrated across 2,500+ repositories, utilizing intelligent agents for iterative verification to automate 60,000+ code reviews per month.
Engineered a robust evaluation framework by building a dependency graph to map C/C++ function hierarchies, enabling AI agents to trace the ripple effects of complex code changes before generating architectural feedback.
Developed a custom RAG architecture utilizing Tree-sitter for AST parsing and syntax verification, alongside vector retrieval to ground LLM reasoning in highly technical spec documents and team-specific design guidelines.
Scaled the ML infrastructure to reliably support this enterprise-grade load, earning selection to present the architecture to the CTO at Qualcomm TechFest (sparQ).
Implemented identity disclosure protocols within the 4G NAS layer to detect and block rogue base station attacks.
Developed "SlideGen," a generative data pipeline utilizing Python APIs to transform raw technical telemetry into structured, compliant corporate formats.

Generative AIPython (Programming Language)Agentic WorkflowsMLOpsRetrieval-Augmented Generation (RAG)AI/ML Infra

Associate Software Engineer

Jul 2024 – Nov 2025 · 1 yr 4 mos · On-site

JavaScriptLarge Language Models (LLM)Python (Programming Language)4GRancher Kubernetes ManagementKubernetes+5

Amd

Intern

Jan 2024 – Jun 2024 · 5 mos · Bengaluru · On-site

Designed and deployed highly optimized, multi-threaded Python scripts utilizing LSF to build distributed data pipelines, analyzing massive hardware bus tracing datasets (250GB+).
Engineered load-balancing solutions that slashed data processing runtime from 7 days to 45 minutes, resolving severe I/O bottlenecks and ensuring peak pipeline efficiency.
Automated complex project management pipelines using JIRA REST APIs, aggregating and hierarchically linking 1,000+ tracking issues to improve system observability.
Developed automated benchmarking infrastructure to evaluate bare-metal server performance (Sysbench, SPEC 2017), directly influencing multi-functional hardware architectural decisions.

Python (Programming Language)Distributed SystemsHigh Performance Computing (HPC)Data PipelinesMultithreading

Kalipatnapu ar/vr lab

President

May 2023 – Jun 2024 · 1 yr 1 mo

Object Oriented DesignLeadershipAugmented Reality (AR)UnityVirtual Reality (VR)Software Development

Standard chartered bank

Summer Intern

May 2023 – Jul 2023 · 2 mos · Chennai, Tamil Nadu, India · On-site

Azure DevOps ServicesMockitoJUnit

Technische universität braunschweig

Technical Researcher

Sep 2022 – Jan 2023 · 4 mos · BITS Pilani - TU Braunschweig · On-site

Participated in Indo-German Sustainability Challenge 2023 with cross-cultural teams from BITS Pilani and TU Braunschweig.
Worked on LCA of E-bike components and identified fields of action for visualization of environmental-related and product data.
Developed a concept for MR-based techniques to support process decisions.
Designed and built a prototype for MR-based visualization tool to aid in process decision-making.

Object Oriented DesignMixed RealityUnityMicrosoft HoloLensSoftware DevelopmentC#

Invoxel technologies

Summer Intern

Jun 2022 – Jul 2022 · 1 mo · Delhi, India

Researched Augmented Reality from a technical and feasibility stand point.
Developed Commercial Augmented Reality Applications in unity
Worked with 3d Models and AR foundation for image, plane recognition.
Used Niantic Lightship ARDK to make virtual elements interact with physical surroundings. Created private Wayspot Anchors to place virtual objects that maintain a consistent, stable pose in an AR environment.

Object Oriented DesignAR FoundationAugmented Reality (AR)UnitySoftware DevelopmentLightship

Iworkslab

Summer Intern

Jun 2022 – Jul 2022 · 1 mo · Hyderabad, Telangana, India

Developed a commercial cross platform e-commerce application that sells non-vegetarian goods using "flutter".
Supervised the App Development team, which created apps for iOS and Android.
Used multiple providers from the provider package to manage the state of the application.
Interacted with the UI team to ensure that the application's flow is simple and easy to use.

Object Oriented DesignAndroid DevelopmentFlutterSoftware Development