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

I am an AI/ML Engineer with 4+ years of experience designing and deploying production-scale Machine Learning and Generative AI systems. I specialize in Retrieval-Augmented Generation (RAG), multi-model LLM orchestration, Agentic AI, and NLP applications using GPT-4-class models, Claude, Llama, LangChain, and LangGraph. I build low-latency AI platforms on PyTorch, Hugging Face, FAISS, Pinecone, Spark, Kafka, and AWS, while following MLOps practices with Docker, Kubernetes, MLflow, and CI/CD to deliver scalable and reliable AI solutions.…I am an AI/ML Engineer with 4+ years of experience designing and deploying production-scale Machine Learning and Generative AI systems. I specialize in Retrieval-Augmented Generation (RAG), multi-model LLM orchestration, Agentic AI, and NLP applications using GPT-4-class models, Claude, Llama, LangChain, and LangGraph. I build low-latency AI platforms on PyTorch, Hugging Face, FAISS, Pinecone, Spark, Kafka, and AWS, while following MLOps practices with Docker, Kubernetes, MLflow, and CI/CD to deliver scalable and reliable AI solutions.

Manikanteswar Gandrothula

Data Scientist, AI Engineer, Full Stack Developer, +2





I am an AI/ML Engineer with 4+ years of experience designing and deploying production-scale Machine Learning and Generative AI systems. I specialize in Retrieval-Augmented Generation (RAG), multi-model LLM orchestration, Agentic AI, and NLP applications using GPT-4-class models, Claude, Llama, LangChain, and LangGraph. I build low-latency AI platforms on PyTorch, Hugging Face, FAISS, Pinecone, Spark, Kafka, and AWS, while following MLOps practices with Docker, Kubernetes, MLflow, and CI/CD to deliver scalable and reliable AI solutions.…I am an AI/ML Engineer with 4+ years of experience designing and deploying production-scale Machine Learning and Generative AI systems. I specialize in Retrieval-Augmented Generation (RAG), multi-model LLM orchestration, Agentic AI, and NLP applications using GPT-4-class models, Claude, Llama, LangChain, and LangGraph. I build low-latency AI platforms on PyTorch, Hugging Face, FAISS, Pinecone, Spark, Kafka, and AWS, while following MLOps practices with Docker, Kubernetes, MLflow, and CI/CD to deliver scalable and reliable AI solutions.

Available to hire

I am an AI/ML Engineer with 4+ years of experience designing and deploying production-scale Machine Learning and Generative AI systems.

I specialize in Retrieval-Augmented Generation (RAG), multi-model LLM orchestration, Agentic AI, and NLP applications using GPT-4-class models, Claude, Llama, LangChain, and LangGraph. I build low-latency AI platforms on PyTorch, Hugging Face, FAISS, Pinecone, Spark, Kafka, and AWS, while following MLOps practices with Docker, Kubernetes, MLflow, and CI/CD to deliver scalable and reliable AI solutions.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Work Experience

AI/ML Engineer at Goldman Sachs

September 1, 2024 - Present

Architected a production-scale Retrieval-Augmented Generation (RAG) framework using embeddings, FAISS, Pinecone, hybrid retrieval, semantic search, and vector indexing techniques to support 200K+ retrieval requests across financial intelligence applications. Developed a multi-model LLM orchestration layer integrating GPT-4-class models, Claude, and Llama-based models with LangChain, dynamic routing, and prompt optimization strategies, reducing inference costs by 27% while maintaining response quality and latency requirements. Optimized latency-sensitive inference pipelines using Redis caching, asynchronous retrieval, embedding precomputation, batching, and concurrent execution techniques, achieving sub-1.8-second p95 latency under 500+ concurrent requests. Fine-tuned transformer-based models using PyTorch, Hugging Face, and PEFT techniques including LoRA adapters to enhance domain-specific understanding and optimize inference efficiency for internal AI applications. Established LLM eva

Machine Learning Engineer at Zensar Technologies

March 1, 2021 - August 31, 2023

Developed a document classification system for insurance and financial documents using TF-IDF, Logistic Regression, SVM, and Random Forest, upgraded to BERT-based transformers (PyTorch, Hugging Face), improving accuracy from 82% to 94% and enabling automated processing. Designed and trained supervised learning models (classification + multi-label tagging) and applied unsupervised clustering (K-Means, PCA-based dimensionality reduction) to group unstructured documents, reducing manual labeling effort by 40%. Configured BERT-based Named Entity Recognition (NER) for extracting key entities such as policy numbers, invoice IDs, and contract clauses, achieving 91% F1-score and improving downstream compliance automation workflows. Implemented distributed data processing pipelines using Apache Spark, handling 3M+ enterprise documents/month, enabling parallel preprocessing, tokenization, and feature extraction for large-scale NLP workloads. Built end-to-end ETL pipelines using Apache Airflow, a