Nov 2023 – Jun 2025 · 1 yr 7 mos

• Developed and deployed a Graph RAG solution utilizing fine-tuned Llama and Neo4J to minimize erroneous hardware dispatches caused by insufficient information during support calls. This event-driven Generative AI agent collects real-time telemetry from the system for self-healing, reducing erroneous dispatches by 60%.
• Improved response quality by using RLHF combined with RLAIF.
• Deployed Llama2-7b-Chat: Implemented Llama2-7b-Chat on premises for Digital Human using NVIDIA TensorRT LLM and Triton inference server on a 4xH100 GPU setup, utilizing pipeline parallelism for 100 parallel customers.
• Optimized Inference: Enhanced Llama2-7b-Chat with paged KV Cache, iterative batching, and context FMHA. Advanced quantization with INT8 AWQ achieved a first token latency of 0.2 seconds for 30 concurrent users.
• Fine-Tuned Llama2-7b-Chat: Utilized QLoRA to fine-tune Llama2-7b-Chat with unstructured data by creating a synthetic dataset, training the model to perform as a customer care representative.
• Led Services Modernization for Dell: Developed a knowledge graph by integrating data from structured and unstructured sources and built a RAG pipeline to provide recommendations to customer representatives.
• Developed Services Modernization Roadmap: Created a comprehensive playbook outlining a 3-5 year roadmap for services modernization, including knowledge graph and LLM based recommendations for root cause analysis and automatic issue resolution using event-based telemetry.

Apr 2021 – Nov 2023 · 2 yrs 7 mos

• Managed the technical execution of Dell partner portal for ITDM with MS Intune integration. Implemented peripheral management for Seamless Ecosystem with expected revenue increase of $3 Million for CSG.
• Implemented Generative AI into the sales process in Dell.com transforming customer buying experience increasing customer satisfaction by 15%.
• Defines and architected battery optimization and browser optimization as part of preventive and self-healing feeding into Dell Optimizer

Jul 2019 – Apr 2021 · 1 yr 9 mos

• Leading Response Time Optimization for commercial platforms and cloud services using AI and ML to ML hinting which improved application performance by 15-20%.
• Introduced design thinking research using product development lifecycle to create products that are customer and design centric to identify customer pain points and create innovation roadmap.
• Defined and executed adding intelligence to Experience management (DEX) ITDM platform for device health increasing product sales by 10% in 6 months

Oct 2022 – Jan 2023 · 3 mos

• Assisting new entrepreneurs for designing their startup and guiding them towards design thinking approach

Jan 2019 – May 2019 · 4 mos · Austin, Texas Area

• Teaching Assistant at LBJ school for Statistical Analysis and Learning.
• Assisting students in data preprocessing and implementing machine learning models

Aug 2018 – Dec 2018 · 4 mos · Austin, Texas Area

• Teaching Assistant at McCombs Business school for nltk and data analytics by scraping websites

Jan 2018 – May 2018 · 4 mos

• Designed and analyzed crowdsourcing task on AMT for collecting fish species, dimensions and polarization angle.
• Performed exploratory data analysis on the polarization dataset to derive the relationship between exploratory and response variables using linear regression in R.
• Created 3D mapping using Point Cloud Library and ORB SLAM2 and GUI in Visual Studio using C++

Aug 2017 – Dec 2017 · 4 mos

• The Transition study for a successful, sustainable software project: Coded an application in python to extract software mentions from research papers using Data Mining.

Jan 2019 – May 2019 · 4 mos · Austin, Texas Area

• Developing software and tools for programming AI network layers in flash, doing inference calculations within the IC, and performing ADC performance analysis using Python and SQL.
• Developing database recorder in SQL for data management and analysis

May 2018 – Jul 2018 · 2 mos · Austin, Texas Area · On-site

• Tested the new technologies of API description languages that allow machines to dynamically read and consume them and to test their state of readiness.
• Created a web application in NodeJS/express and Angular that dynamically discovers consolidates the services of Microsoft Intune and VMWare Airwatch into a unified console and incorporates updates automatically.

Mar 2014 – Dec 2016 · 2 yrs 9 mos · Jalandhar Area, India

• Designed the user interface of an apparels website, “Oradeals.com”, using Java in Eclipse IDE and also implemented its backend in SQL 2008.
• Coded physical database of an interior designer website, “Gagan Architect” in Oracle 10g. Responsible for integration, functional and acceptance testing of the website. Also designed test suites for regression testing at all development levels.

Jun 2013 – Feb 2014 · 8 mos · Noida Area, India

• Designed and validated transceiver module. Responsible for coding C-based transceivers and CRC modules. Implemented and tested drivers using cryptographic operations to facilitate BootRom security.

Jan 2012 – Jul 2012 · 6 mos · Gurgaon, India

• Designed testcases and test-suites for testing different flows in Eclipse IDE. Also responsible for regression testing and unit testing of Java based applications using Selenium IDE and Test NG.
• Designed, coded and tested “Chat Client” in Core Java, encompassing features like message broadcasting.