Senior ML Engineer - LLMs, RAG, MLOps, NLP Leadership

Vedika Agarwal

+1-470-***-**** # ***********@******.*** ï /vedikagarwal vedikagarwal.github.io Atlanta, GA SUMMARY

Georgia Tech Machine Learning Graduate (4.0 GPA) and Senior Engineer with 3+ years at HPE building production AI systems. Delivered 90–95% gains for 65K+ customers via RAG pipelines, LLM fine-tuning & end-to-end automation. NASA-funded ML researcher. EDUCATION

Georgia Institute of Technology Aug 2024 – May 2026 M.S. in Computer Science, GPA: 4.0/4.0 Atlanta, GA

• Specialization in Machine Learning

• Coursework: Machine Learning, Deep Learning, Natural Language Processing, Vision Language Models, Advanced Algorithms R.V. College of Engineering Aug 2017 – Jul 2021

B.E. in Computer Science Bangalore, IN

• Coursework: Distributed Programming, Computer Architecture, Operating Systems, Computer Networks EXPERIENCE

NASA funded iCooL Lab – Georgia Institute of Technology Jan 2025 – Present Graduate Research Assistant Atlanta, GA

• Built a multi-stage ML pipeline (Random Forest, LSTM, kNN) for long-term chemical evolution simulation, reducing discovery timelines from years to <4 hours of inference; selected to present at 2026 NASA AbSciCon. Hewlett Packard Enterprise Sep 2021 – Jul 2024

Senior Software Engineer Bangalore, IN

• Led a 4-engineer team to build an AI powered chatbot for cloud infrastructure; recognised by HPE CTO office.

• Fine-tuned GPT-2 and implemented RAG on OCR and Confluence data for troubleshooting, reducing resolution time by 90%.

• Reduced platform bring-up by 95% (72h to <4h) for 65K+ enterprise customers by engineering end-to-end automation.

• Published IEEE paper “Benchmarking LLMs for Zero-Day Vulnerabilities”; identifying security issues by fine tuning LLMs.

• Built pre/post-deployment validation pipeline with Apache Airflow DAGs, cutting customer shipment from weeks to hours.

• Engineered network security automation tool, 32+ ACL rules across 5 VLANs, reducing manual effort 94% (35 min to 2 min). Hewlett Packard Enterprise Feb 2021 – Aug 2021

Software Engineer Intern Bangalore, IN

• Built end to end OpenShift ISV solution, reducing platform bring-up from 1 week to 3 hrs; automating OS deployment, core services

• Improved IPV4/IPV6 dual-stack provisioning for HPE OneView (1M+ customers), 0% regression, reducing product shipment time. PROJECTS

VPTmed: DINOv3-Powered Spatial Reasoning for VLMs in Medical Imaging Report Aug 2025 – Dec 2025

• Architected a dual-encoder VLM (Qwen2-VL + DINOv2-ViT-H) with Visual Prompt Tuning and FreeFreeze fine-tuning.

• Achieved 27.2% exact-match accuracy on VQA-Med 2019, surpassing BLIP MedVQA baseline by +8.4pp. Drug-Target Interaction Prediction GitHub May 2025 – Aug 2025

• Built end-to-end ML pipeline predicting AChE inhibitor bioactivity across 27K+ compounds from ChEMBL.

• Extracted molecular descriptors via RDKit and benchmarked RF, SVM, and neural networks for IC50 prediction. Human Motion Retargeting for Humanoid Robots GitHub Jan 2025 – May 2025

• Built video-to-robot motion pipeline using HybrIK for 3D pose extraction, trajectory optimization, and PPO-based RL in MuJoCo.

• TO-based policy achieved 0.021 rad joint tracking error and 1.58cm hand accuracy, outperforming IK baseline across all metrics. TECHNICAL SKILLS

Languages & Certifications : Python, C/C++, SQL, Bash, Go, AWS Certified Cloud Practitioner (CLF-C02) ML & AI : PyTorch, TensorFlow, Scikit-learn, NLTK, CNNs, Transformers, Reinforcement Learning Systems & Cloud : Docker, Kubernetes, Terraform, Ansible, Airflow, Jenkins, Linux Data & Tools : NumPy, Pandas, SciPy, Git, REST APIs Achievements

• Awarded 100% merit scholarship, dual-degree M.S. at Georgia Tech (top 2% applicants), research and academic excellence.

• Won Highest Innovation Award at HPE, placing in top 1% of engineers by delivering award-winning projects at HPE TechCon.

• Supported academic success of 300+ undergraduates as Georgia Tech GTA, leading technical review sessions and managing grading.

Contact this candidate