Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I'm Shamil Lathiya, an AI/ML Engineer specializing in Generative AI and MLOps. I design and deploy LLM-powered solutions, RAG systems, and multi-agent workflows for enterprise environments. With experience at ServiceNow and Observe.AI, I help teams accelerate release cycles, improve model quality, and automate complex processes through end-to-end MLOps, scalable deployments, and low-latency inference.…I'm Shamil Lathiya, an AI/ML Engineer specializing in Generative AI and MLOps. I design and deploy LLM-powered solutions, RAG systems, and multi-agent workflows for enterprise environments. With experience at ServiceNow and Observe.AI, I help teams accelerate release cycles, improve model quality, and automate complex processes through end-to-end MLOps, scalable deployments, and low-latency inference.

Shamil Lathiya

AI Engineer, Data Scientist, Web Developer, +2





I'm Shamil Lathiya, an AI/ML Engineer specializing in Generative AI and MLOps. I design and deploy LLM-powered solutions, RAG systems, and multi-agent workflows for enterprise environments. With experience at ServiceNow and Observe.AI, I help teams accelerate release cycles, improve model quality, and automate complex processes through end-to-end MLOps, scalable deployments, and low-latency inference.…I'm Shamil Lathiya, an AI/ML Engineer specializing in Generative AI and MLOps. I design and deploy LLM-powered solutions, RAG systems, and multi-agent workflows for enterprise environments. With experience at ServiceNow and Observe.AI, I help teams accelerate release cycles, improve model quality, and automate complex processes through end-to-end MLOps, scalable deployments, and low-latency inference.

Available to hire

I’m Shamil Lathiya, an AI/ML Engineer specializing in Generative AI and MLOps. I design and deploy LLM-powered solutions, RAG systems, and multi-agent workflows for enterprise environments.

With experience at ServiceNow and Observe.AI, I help teams accelerate release cycles, improve model quality, and automate complex processes through end-to-end MLOps, scalable deployments, and low-latency inference.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Work Experience

AI/ML Engineer at ServiceNow

July 1, 2024 - Present

Fine-tuned instruction-tuned LLMs using LoRA/QLoRA and Hugging Face Transformers, improving model accuracy by 20% and reducing training costs by 18% compared to full fine-tuning. Built and deployed RAG pipelines with LangChain and Pinecone to improve contextual response precision by 25% and reduce query latency in enterprise knowledge retrieval. Developed Agentic AI workflows and multi-agent systems using LangGraph, function calling, and structured outputs to automate complex multi-step enterprise processes and cut manual interventions by 30%. Created an ML-based Incident Auto-Classification model that reduced manual ticket routing time by 60% and improved categorization accuracy across ITSM pipelines. Implemented a Predictive Maintenance system using LSTM time-series analysis, integrated with ServiceNow Event Management to detect infrastructure failures before occurrence, reducing unplanned downtime by 22%. Orchestrated end-to-end MLOps pipelines with MLflow, Kubernetes, and CI/CD on

AI/ML Engineer at Observe.AI

January 1, 2021 - June 1, 2022

Developed NLP and Conversational AI solutions using Python and Transformer-based models, improving customer interaction quality and automating support workflows via FastAPI services. Built Intent Classification and Named Entity Recognition (NER) pipelines using PyTorch and Scikit-learn, improving query understanding accuracy by 25% and reducing manual handling across customer conversations. Designed data preprocessing, feature engineering, and semantic matching workflows to improve response relevance and consistency across support automation systems. Deployed and maintained ML services using Docker, AWS (S3, Lambda), and CI/CD pipelines, ensuring reliable inference and stable production releases with PostgreSQL and MongoDB. Monitored model performance, analyzed inference metrics, and optimized latency by 20% to improve system efficiency and user experience.