ARUN PALANISAMY

I am a Machine Learning Engineer with 3+ years of hands-on experience designing and deploying NLP and document intelligence solutions for finance automation. I have built end-to-end AI pipelines that include OCR, RAG systems, forecasting models, and multi-agent orchestration to drive operational efficiency.

In production, I have fine-tuned LLMs (GPT variants, BERT, HuggingFace), integrated retrieval with FAISS and LangChain, and deployed microservices using FastAPI in Docker on Azure. My work emphasizes scalable, monitored systems with automated retraining to maintain long-term accuracy.

At Paperchase Accountancy I led projects that reduced manual entry and review times substantially — for example, OCR-driven extraction improved accuracy and cut manual effort by 65%, while an LLM-based financial summarizer reduced analyst review time by 40%. I also implemented multi-agent AI for compliance, reducing processing time by 60%.

I bring strong engineering discipline—Python, PyTorch/TensorFlow, SQL, CI/CD, Kubernetes—and a practical mindset focused on measurable business impact, reliable deployments, and continual improvement through monitoring and automation.