Tejashwini Reddy

About Tejashwini Reddy

I’m a Senior GenAI & AI/ML Engineer with 10+ years of experience building large-scale AI systems across finance, healthcare, retail, and enterprise analytics. I specialize in LLM fine-tuning, RAG architectures, multimodal ML, GPU optimization, and end-to-end MLOps pipelines.I’ve built high-impact AI platforms for clients like Goldman Sachs, Boston Scientific, Spectrum Reach, and State of Ohio—covering fraud detection, medical imaging, personalized advertising, and public health analytics. My work spans LLMs (GPT, LLaMA, Claude), Transformers, LangChain, Pinecone, LoRA/QLoRA, multimodal CNNs, distributed training, and real-time inference on Kubernetes with 99.9%+ uptime.I’m passionate about turning complex problems into production-ready AI solutions that are scalable, secure, and cost-efficient. I bring deep experience in AWS, Azure, MLflow, Triton, Airflow, Data Engineering, and compliance (PCI/HIPAA). I thrive in environments that value innovation, speed, and measurable business impact.

Experience

1. Gen Ai Engineer

   Goldman Sachs

   Oct 2024 — Present · New York, NY, US

   At Goldman Sachs, I lead the development of next-generation GenAI fraud detection systems using RAG, LLM fine-tuning, and real-time streaming pipelines. I boosted fraud detection accuracy and recall through GPT-4/LLaMA fine-tuning (LoRA/QLoRA), reduced false positives by 85%, and deployed ultra-low-latency models (<20 ms) across global EKS clusters. I engineered scalable data pipelines on Kafka/Kinesis, built vector search with Pinecone/FAISS for coordinated fraud detection, and automated CI/CD for compliant LLM releases—cutting model update cycles from weeks to days. My work directly strengthens AML/KYC compliance, reduces operational cost, and increases the speed and reliability of enterprise fraud decisions.

Education

• Vellore Institute of Technology

  Bachelor of Technology - BTech, Computer Science