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

I'm Viyani Sushmitha Maria Joseph, an AI Engineer with 7+ years of experience building autonomous AI agents, multi-agent systems, and machine learning solutions across software engineering, fintech, and ecommerce. I specialize in Retrieval-Augmented Generation (RAG), Semantic Search, reasoning-driven AI workflows, and real-time inference systems, with hands-on work using Python, PySpark, TensorFlow, Scikit-learn, LangGraph, and LangChain. I have deployed scalable AWS-based AI infrastructure and built LLM evaluation and observability frameworks to ensure reliable production AI systems that support automation and decision-making. In roles at Cognition AI and PhonePe, I designed and implemented autonomous agent orchestration with tool-calling, memory management, and contextual planning to improve workflow accuracy and efficiency. I built RAG pipelines with Pinecone and FastAPI, optimized real-time LLM inference on AWS EKS, and established evaluation pipelines with LangSmith and DeepEval to accelerate issue resolution. I thrive in cross-functional teams and enjoy delivering practical AI solutions that scale.…I'm Viyani Sushmitha Maria Joseph, an AI Engineer with 7+ years of experience building autonomous AI agents, multi-agent systems, and machine learning solutions across software engineering, fintech, and ecommerce. I specialize in Retrieval-Augmented Generation (RAG), Semantic Search, reasoning-driven AI workflows, and real-time inference systems, with hands-on work using Python, PySpark, TensorFlow, Scikit-learn, LangGraph, and LangChain. I have deployed scalable AWS-based AI infrastructure and built LLM evaluation and observability frameworks to ensure reliable production AI systems that support automation and decision-making. In roles at Cognition AI and PhonePe, I designed and implemented autonomous agent orchestration with tool-calling, memory management, and contextual planning to improve workflow accuracy and efficiency. I built RAG pipelines with Pinecone and FastAPI, optimized real-time LLM inference on AWS EKS, and established evaluation pipelines with LangSmith and DeepEval to accelerate issue resolution. I thrive in cross-functional teams and enjoy delivering practical AI solutions that scale.

Viyani Sushmitha Maria Joseph

AI Engineer, Data Scientist, Full Stack Developer, +1





I'm Viyani Sushmitha Maria Joseph, an AI Engineer with 7+ years of experience building autonomous AI agents, multi-agent systems, and machine learning solutions across software engineering, fintech, and ecommerce. I specialize in Retrieval-Augmented Generation (RAG), Semantic Search, reasoning-driven AI workflows, and real-time inference systems, with hands-on work using Python, PySpark, TensorFlow, Scikit-learn, LangGraph, and LangChain. I have deployed scalable AWS-based AI infrastructure and built LLM evaluation and observability frameworks to ensure reliable production AI systems that support automation and decision-making. In roles at Cognition AI and PhonePe, I designed and implemented autonomous agent orchestration with tool-calling, memory management, and contextual planning to improve workflow accuracy and efficiency. I built RAG pipelines with Pinecone and FastAPI, optimized real-time LLM inference on AWS EKS, and established evaluation pipelines with LangSmith and DeepEval to accelerate issue resolution. I thrive in cross-functional teams and enjoy delivering practical AI solutions that scale.…I'm Viyani Sushmitha Maria Joseph, an AI Engineer with 7+ years of experience building autonomous AI agents, multi-agent systems, and machine learning solutions across software engineering, fintech, and ecommerce. I specialize in Retrieval-Augmented Generation (RAG), Semantic Search, reasoning-driven AI workflows, and real-time inference systems, with hands-on work using Python, PySpark, TensorFlow, Scikit-learn, LangGraph, and LangChain. I have deployed scalable AWS-based AI infrastructure and built LLM evaluation and observability frameworks to ensure reliable production AI systems that support automation and decision-making. In roles at Cognition AI and PhonePe, I designed and implemented autonomous agent orchestration with tool-calling, memory management, and contextual planning to improve workflow accuracy and efficiency. I built RAG pipelines with Pinecone and FastAPI, optimized real-time LLM inference on AWS EKS, and established evaluation pipelines with LangSmith and DeepEval to accelerate issue resolution. I thrive in cross-functional teams and enjoy delivering practical AI solutions that scale.

Available to hire

I’m Viyani Sushmitha Maria Joseph, an AI Engineer with 7+ years of experience building autonomous AI agents, multi-agent systems, and machine learning solutions across software engineering, fintech, and ecommerce. I specialize in Retrieval-Augmented Generation (RAG), Semantic Search, reasoning-driven AI workflows, and real-time inference systems, with hands-on work using Python, PySpark, TensorFlow, Scikit-learn, LangGraph, and LangChain. I have deployed scalable AWS-based AI infrastructure and built LLM evaluation and observability frameworks to ensure reliable production AI systems that support automation and decision-making.

In roles at Cognition AI and PhonePe, I designed and implemented autonomous agent orchestration with tool-calling, memory management, and contextual planning to improve workflow accuracy and efficiency. I built RAG pipelines with Pinecone and FastAPI, optimized real-time LLM inference on AWS EKS, and established evaluation pipelines with LangSmith and DeepEval to accelerate issue resolution. I thrive in cross-functional teams and enjoy delivering practical AI solutions that scale.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Work Experience

AI Engineer at Cognition AI

July 1, 2025 - Present

Developed autonomous AI Agents using LangGraph, LangChain, and LLMs, enabling multi-step reasoning and task execution workflows across enterprise applications. Architected Multi-Agent Systems with tool-calling, memory management, and contextual planning capabilities, increasing complex workflow completion accuracy for production AI environments. Built Retrieval-Augmented Generation (RAG) pipelines using Pinecone, Semantic Search, and FastAPI, improving knowledge retrieval relevance across large-scale datasets. Optimized real-time LLM inference infrastructure on AWS EKS with Docker, Kubernetes, and Redis, reducing response latency while supporting high concurrency. Established LLM evaluation and observability frameworks using LangSmith, DeepEval, and Ragas, increasing validation coverage and accelerating issue resolution. Automated deployment and monitoring workflows through GitHub Actions, Prometheus, and Grafana, improving production reliability across distributed AI services.

AI Engineer at PhonePe

May 1, 2020 - November 1, 2023

Developed Fraud Detection Models using PySpark and XGBoost pipelines, reducing suspicious transaction investigation turnaround time for risk analysis teams. Built intelligent retrieval workflows with multilingual support APIs to enhance contextual responses for merchant-facing applications. Evaluated Recommendation Systems using TensorFlow experimentation workflows, improving personalized financial product engagement among regional users. Implemented NLP inference APIs through FastAPI and Docker, enabling real-time multilingual payment assistance. Streamlined real-time payment analytics using Apache Spark and Airflow, reducing delayed fraud alert generation during peak windows. Strengthened production observability through Prometheus and Grafana dashboards, improving monitoring response times for inference services on AWS Lambda infrastructure.

ML Engineer at CommerceIQ

February 1, 2017 - March 1, 2020

Developed Forecasting Models using Scikit-learn on retail datasets to improve inventory planning accuracy. Refined Recommendation Systems algorithms through PyTorch experimentation workflows, improving personalized product ranking quality across digital commerce merchandising environments. Processed large-scale retail datasets using Databricks and PySpark transformations, reducing reporting inconsistencies across analytics pipelines. Automated ETL Pipelines using Python services and Airflow scheduling workflows, supporting reliable pricing intelligence and inventory optimization model training. Validated machine learning feature quality through statistical evaluation processes, enhancing downstream model consistency. Deployed forecasting inference services using AWS SageMaker and REST APIs, enabling scalable product demand predictions for ecommerce revenue optimization.