Ai Engineer Cloud

Jothiswaran Arumugam

AI Engineer - *.*+ YOE in GenAI, RAG, Cloud MLOps and Multi-Agent Systems

Tallahassee, USA 850-***-**** ********.***********@*****.*** LinkedIn Github Portfolio WORK EXPERIENCE

AI Engineer (Intern & Co-op) May 2024 - May 2025

IM Digital Boca Raton, FL

• Delivered sub-300ms retrieval-augmented generation for legal and product domains via semantic search and hybrid vector indexing.

• Fine-tuned LLaMA & Mistral via LoRA on Vertex Pipelines, improving generation accuracy by 37% and lowering token cost by 35%.

• Shipped multi-agent LLM APIs with FastAPI and Cloud Run, handling 1K+ daily queries with RBAC and end-to-end observability.

• Streamlined clause analysis and product Q&A using orchestrated RAG agents, reducing human review workload by 40%.

• Enabled CI/CD-driven LLM evaluation via LangSmith + RAGAS, improving hallucination grounding scores by 62% in production. GCP Cloud Engineer Jul 2021 - Aug 2023

CTS Hyderabad, TG

• Automated ML infrastructure provisioning using Terraform, GKE, and BigQuery, reducing setup time by 80% across 15+ projects.

• Deployed real-time LLM inference pipelines via Vertex AI and CI/CD, improving latency observability and SLA adherence by 60%.

• Optimized ETL workflows in Dataflow and BigQuery, cutting cloud compute costs by 30% and boosting daily pipeline throughput.

• Mentored 30+ engineers in GCP, serverless ML, and AI deployment best practices, accelerating internal adoption by 4x. Machine Learning Intern May 2020 - Jul 2020

Verzeo Hyderabad, TG

• Developed CNNs in Keras for pneumonia detection from X-rays, achieving 82% accuracy after OpenCV-based preprocessing.

• Developed supervised NLP pipelines for sentiment analysis and classification, improving F1 scores by 25% post tuning.

• Tuned DL models using TensorFlow and PyTorch, accelerating training cycles by 40% via optimized hyperparameter sweeps. EDUCATION

Florida State University Aug 2023 - May 2025

Master of Science in Computer Science, GPA: 3.72/4.0 Tallahassee, FL Rajalakshmi Engineering College Aug 2017 - May 2021 Bachelor of Engineering in Computer Science, GPA: 3.5/4 Chennai, TN TECHNICAL SKILLS

LLMs & GenAI: OpenAI API, Claude, Gemini, HuggingFace, LLaMA, RAG, Prompt Engineering, LoRA, PEFT, Function Calling ML & NLP: TensorFlow, PyTorch, Scikit-learn, AutoML, Transfer Learning, CNNs, Text Classification, NER MLOps & Deployment: Vertex AI, Kubeflow, MLflow, CI/CD, Docker, Kubernetes, FastAPI, Airflow, Cloud Run Cloud & Infra: GCP, AWS, Lambda, Kafka, Telnyx SIP, LiveKit, VPC, Serverless APIs, Pub/Sub Data & Retrieval: Pinecone, FAISS, Qdrant, Weaviate, BigQuery, Dataflow, Pandas, NumPy, PostgreSQL, MongoDB Security & Evaluation: RBAC, PII Redaction, Responsible AI, LangSmith, RAGAS, Prometheus, Grafana Languages & Tools: Python, Java, SQL, PostgreSQL, MongoDB, TypeScript, React.js, Git, Bash, Linux CERTIFICATIONS (Google Cloud Certified 7x)

• GCP - Professional Machine Learning Engineer

• GCP - Professional Cloud Architect

• GCP - Professional Data Engineer

• GCP - Professional Cloud Developer

• GCP - Professional Cloud DevOps Engineer

• GCP - Professional Cloud Database Engineer

• GCP - Associate Cloud Engineer

PROJECTS

VoiceFlow AI – Conversational Agent Deepgram STT, Groq LLM, Google TTS, Telnyx, LiveKit Apr 2025 - May 2025

• Implemented a 24/7 real-time voice agent with Whisper, Deepgram, Groq, and Google TTS, achieving sub-800ms end-to-end latency.

• Integrated Telnyx SIP + LiveKit audio streaming with LangChain RAG and local vector DB for low-latency semantic response. Financial Crew – Real-Time Stock Analyst CrewAI, Sambanova LLM, YFinance, Streamlit, Python Mar 2025 - Apr 2025

• Designed and deployed CrewAI agents using SambaNova LLM for stock QA, trend insights, and markdown report generation.

• Integrated YFinance and Streamlit into tool-using agents with modular prompt chains, tested on 15+ real stock tickers. LegalAI – Multi-Agent System for Legal Risk Analysis LangGraph, Transformer, OpenAI, Docker Feb 2025 - Apr 2025

• Created LangGraph LLM agents for clause extraction, contradiction detection, and compliance review; cut review time by 60%.

• Scaled Gemini-backed agents with memory and task routing across 10+ structured legal document types. ArXiv Explorer – A Modular AI Research Assistant LangChain, Qdrant, Gemini, Streamlit Jan 2025 - Feb 2025

• Engineered RAG system with LangChain, Gemini, and Qdrant to semantically index and summarize 25K+ ArXiv papers.

• Built Streamlit UI for topic-based paper retrieval, live summarization, and real-time NLP exploration.

Contact this candidate