Home

Rithvik Reddy - AI/ML Engineer

[email protected]

Location: Texas City, Texas, USA

Relocation: Yes

Visa: H1B Visa

Resume file: Rithvik Pinninti Resume_1774984284435.pdf Please check the file(s) for viruses. Files are checked manually and then made available for download.

RITHVIK REDDY PINNINTI [email protected] | (806) 702-3056 6+ Years | Senior AI/ML Engineer | LLM / GenAI / MLOps / Cloud-Native ML Platforms CAREER OBJECTIVE Senior Machine Learning Engineer with 7+ years of experience building production AI systems and scalable ML infrastructure. Specialized in LLM deployment, GPU-optimized inference, and cloud-native ML platforms using Python, Kubernetes, and distributed computing. Proven track record of operationalizing models that serve high-throughput workloads with strong reliability and performance. Strong expertise in LLM infrastructure, model hosting, fine-tuning (LoRA/QLoRA), RAG pipelines, embeddings, and vector database integrations for real-time production systems. Experienced in taking AI/ML and GenAI solutions from proof-of-concept to production with focus on scalability, reliability, latency optimization, and cost efficiency. Skilled in building cloud-native AI platforms using AWS, GCP, and Azure with Docker, Kubernetes, CI/CD pipelines, and automated model monitoring. Hands-on experience with MLOps, experiment tracking, model registry, drift detection, and continuous retraining pipelines ensuring stable and reliable model performance. PROFESSIONAL SUMMARY Senior Machine Learning Engineer with 6+ years of experience building production AI systems and scalable ML infrastructure across banking, retail, telecom, and insurance industries. Specialized in LLM deployment, GPU-optimized inference, and cloud-native ML platforms using Python, Kubernetes, and distributed computing with sub-second response times for high-volume enterprise workloads. Strong expertise in LLM infrastructure, model hosting, fine-tuning (LoRA/QLoRA), RAG pipelines, embeddings, and vector database integrations (Pinecone, FAISS, Chroma, Weaviate) for real-time production systems. Experienced in Agentic AI workflows using LangChain, AutoGen, CrewAI with ReAct patterns, multi-agent orchestration, tool calling, and memory-augmented agents for intelligent automation. Skilled in building cloud-native AI platforms using AWS (SageMaker, Lambda, EC2, S3, ECS), GCP (Vertex AI, Dataflow, BigQuery), Azure ML with Docker, Kubernetes, CI/CD pipelines, and automated model monitoring. Hands-on experience with MLOps and LLMOps including MLflow, Weights & Biases, experiment tracking, model registry, A/B testing, canary deployments, drift detection, and continuous retraining pipelines. Deep expertise in Deep Learning frameworks including PyTorch, TensorFlow, Keras, ONNX, TensorRT, Detectron2, YOLO, Transformers, mixed precision training, and distributed training. Strong data engineering skills with PySpark, Apache Spark, Apache Beam, Apache Airflow, Kafka, BigQuery, Snowflake, SQL, feature stores, and streaming/batch data processing. Proven ability to collaborate with cross-functional product teams and ML research scientists translating POC experiments into scalable production systems with load testing, security hardening, and compliance pipelines. AI & GENAI CAPABILITIES TECHNOLOGY DEPLOYMENT BUSINESS VALUE LLM Infrastructure & Hosting Production GPU Clusters Sub-Second Inference TECHNICAL DEPTH Hugging Face, Triton Server, PyTorch, TensorRT, LoRA/QLoRA Fine-Tuning, Model Quantization, Tokenization Optimization RAG & Vector Databases Enterprise Production Reduced Hallucinations Pinecone, FAISS, Chroma, Weaviate, Embedding Pipelines, Sentence Transformers, Contextual Retrieval Agentic AI & Orchestration Workflow Automation Intelligent Automation LangChain Agents, AutoGen, CrewAI, ReAct Patterns, Multi-Agent Orchestration, Tool Calling, Memory-Augmented Agents Predictive ML Models 3M+ Daily Predictions Business-Critical Decisions scikit-learn, XGBoost, LightGBM, Ensemble Methods, Classification, Regression, Time Series Forecasting, SHAP/LIME Explainability NLP & Text Analytics Multi-Domain Actionable Insights TF-IDF, Topic Modeling, Sentiment Analysis (VADER), NER, Text Classification, Gensim LDA, SentenceTransformers Computer Vision Real-Time Analytics GPU-Accelerated PyTorch, YOLO, Detectron2, TensorRT, ONNX, Mixed Precision Training, Distributed Training INTEGRATION & MLOPS TECHNOLOGY LEVEL SCOPE TECHNICAL SKILLS MLflow & Model Registry Expert End-to-End MLOps Experiment Tracking, Model Versioning, A/B Testing, Canary Deployments, Blue-Green Releases, Automated Rollback CI/CD for ML Expert Pipeline Automation GitHub Actions, Jenkins, Docker Build, ECR Push, Linting, Unit/Integration Tests, Quality Gates Kubernetes & Containers Expert Orchestration Docker, K8s, StatefulSets, Autoscaling, GPU Scheduling, Terraform, Microservices Architecture Monitoring & Observability Expert Production Reliability Prometheus, Grafana, CloudWatch, Stackdriver, Drift Detection, Evidently AI, Alerting Systems Data Pipelines Advanced Feature Engineering PySpark, Apache Spark, Airflow, Beam, Kafka, BigQuery, Snowflake, Feature Stores, ETL/ELT CORE SKILLS DOMAIN TOOLS & TECHNOLOGIES LLM & Generative/Agentic AI LLM Training, Fine-Tuning (LoRA, QLoRA, PEFT), RAG, Embeddings (Sentence Transformers, OpenAI), Tokenization, GPU Utilization, Inference Optimization, Triton Inference Server, Hugging Face Transformers, Vector Databases (Pinecone, Chroma, FAISS, Weaviate), Prompt Engineering, Model Quantization, LangChain Agents, ReAct, Multi-Agent Orchestration, Tool Calling, Memory-Augmented Agents AI/ML Engineering Regression & Classification Models, NLP (TF-IDF, Topic Modeling, Sentiment Analysis, NER), scikit-learn, VADER, Pandas, NumPy, PySpark ML, Feature Engineering, Time Series Forecasting, Model Explainability (SHAP, LIME), Experiment Tracking, Hyperparameter Tuning, Ensemble Methods, LightGBM, XGBoost MLOps & LLMOps CI/CD for ML, MLflow, Weights & Biases, Model Registry, A/B Testing, Canary Deployments, Blue-Green Deployments, Automated Retraining, Drift Detection, Performance Monitoring, GPU Scaling, Apache Airflow, Apache Beam, ETL Pipelines, Model Versioning, Shadow Testing, Model Risk Oversight Cloud & DevOps AWS (SageMaker, Lambda, EC2, S3, ECS), GCP (Vertex AI, Dataflow, BigQuery), Azure ML, Docker, Kubernetes, Terraform, FastAPI, Flask, Microservices Architecture, Event-Driven Pipelines, CloudWatch, Stackdriver, Prometheus, Grafana Data Engineering PySpark, Apache Spark, Apache Beam, Apache Airflow, Kafka, BigQuery, Snowflake, SQL, Data Warehousing, Streaming Data, Batch Processing, ETL/ELT, Data Quality, Feature Stores Deep Learning PyTorch, TensorFlow, Keras, ONNX, TensorRT, Detectron2, YOLO, Transformers, Mixed Precision Training, Distributed Training PROFESSIONAL EXPERIENCE Project 1: BMO Harris Bank (LLM Infrastructure & Enterprise AI Platform) Role: AI/ML Engineer | Duration: Nov 2024 Present Location: Chicago, IL Environment: Hugging Face Transformers, Triton Inference Server, PyTorch, TensorRT, Docker, Kubernetes, Pinecone, Chroma, FAISS, MLflow, Prometheus, Grafana, FastAPI, AWS SageMaker, LangChain, CrewAI, AutoGen Architected GPU-accelerated LLM inference platform using Hugging Face, Triton, and PyTorch, enabling sub-second response times for high-volume enterprise workloads across BMO s banking operations. Fine-tuned large language models (LLaMA, Mistral) using LoRA/QLoRA techniques, improving domain-specific accuracy while reducing training costs significantly for banking-specific use cases. Designed containerized deployment workflows with Docker and Kubernetes, implementing autoscaling and blue green releases to ensure high availability and zero-downtime deployments. Built scalable RAG pipelines integrating vector databases (Pinecone/FAISS) to enhance contextual retrieval and reduce hallucinations in enterprise knowledge systems. Established CI/CD pipelines for ML models using GitHub Actions and MLflow, accelerating deployment cycles and improving model governance with automated versioning and rollback capabilities. Implemented real-time monitoring with Prometheus and Grafana, tracking latency, throughput, and GPU utilization to proactively detect performance bottlenecks and ensure SLA compliance. Developed secure API-based model serving frameworks with rate limiting, content filtering, and audit logging to meet enterprise compliance standards and regulatory requirements. Developed and optimized AI applications using Python, LangChain, and CrewAI, implementing Retrieval-Augmented Generation (RAG) and Agentic AI workflows for intelligent automation of banking processes. Developed autonomous agent workflows using LangChain and AutoGen to enable intelligent task automation with multi-agent orchestration, tool calling, and memory-augmented agents. Built scalable API services for serving RAG-based generative AI solutions on AWS SageMaker and deployed models for scalable inference and real-time predictions across multiple business units. Built centralized model management platform using MLflow, enabling automated versioning, A/B testing, and rollback for production reliability across all ML models in the organization. Developed detailed technical documentation including API specifications with OpenAPI schemas, architectural diagrams with Lucidchart, performance benchmarking reports, and operational runbooks covering incident response, scaling procedures, and troubleshooting workflows. Collaborated with cross-functional product teams and ML research scientists translating POC experiments into scalable production systems implementing load testing with Locust, security hardening with prompt injection detection, and PII redaction pipelines. Project 2: BestBuy (ML Model Operationalization & Production Deployment) Role: AI/ML Engineer | Duration: Aug 2023 Nov 2024 Location: Richfield, Minnesota Environment: Python, PySpark, scikit-learn, VADER, Pandas, NumPy, TF-IDF, Gensim LDA, SentenceTransformers, MLflow, Apache Airflow, Docker, Kubernetes, FastAPI, Kafka, BigQuery, SHAP, Evidently AI, Databricks Productionized machine learning models using scikit-learn and gradient boosting frameworks, supporting 3+ million daily predictions across business-critical retail applications including demand forecasting and customer analytics. Engineered distributed ETL pipelines using PySpark and Databricks to support large-scale feature engineering and automated retraining workflows processing terabytes of retail transaction data. Built RESTful inference services using FastAPI, enabling real-time and batch prediction capabilities for downstream systems with comprehensive error handling and request validation. Implemented automated model monitoring with drift detection and performance metrics using Evidently AI, improving model reliability and reducing production incidents across the ML fleet. Containerized ML workloads using Docker and deployed to Kubernetes, significantly improving scalability and deployment consistency across development, staging, and production environments. Developed NLP pipelines leveraging TF-IDF, embeddings, and NER to extract actionable insights from unstructured customer reviews and product descriptions at scale. Designed A/B testing frameworks to evaluate model performance with statistical significance testing, enabling data driven release decisions for business-critical ML models. Built automated monitoring pipelines generating model performance metrics including AUC-ROC curves, precision recall analysis, confusion matrices, statistical drift detection using KS tests and PSI calculations, feature importance tracking with SHAP values, and compliance dashboards for Model Risk Oversight teams. Automated end-to-end CI/CD pipelines in Jenkins: linting Python code, running unit and integration tests, building Docker images, and pushing to Amazon ECR with automated quality gates. Collaborated with business stakeholders and data analysts translating requirements into ML solutions implementing custom loss functions, business constraint optimization, and explainability reports using LIME local interpretations ensuring stakeholder confidence. Project 3: Cisco (GPU-Accelerated ML Systems & Computer Vision) Role: Python & ML Engineer | Duration: Aug 2020 Jul 2022 Location: San Jose, California Environment: PyTorch, YOLO, TensorRT, Docker, Kubernetes, MLflow, PySpark, Apache Beam, Kafka, Redis, Prometheus, Grafana, FastAPI, RabbitMQ, Terraform, AWS EC2 Built GPU-accelerated computer vision and ML inference systems using PyTorch, YOLO, and TensorRT for real time analytics across Cisco s enterprise networking and security product lines. Deployed containerized ML pipelines on Kubernetes and AWS improving scalability and reliability of AI workloads with auto-scaling, health checks, and resource optimization. Developed machine learning models for forecasting, classification, and recommendation systems across enterprise projects supporting network traffic analysis and anomaly detection. Built large-scale Spark and Airflow data pipelines for feature engineering and analytics across structured and unstructured datasets with automated scheduling and monitoring. Implemented end-to-end MLOps including model versioning, monitoring, validation, and automated retraining workflows using MLflow for experiment tracking and model registry management. Developed NLP solutions for sentiment analysis, topic modeling, and text classification improving business insights from customer feedback and support ticket analysis. Optimized inference performance using quantization and batching techniques reducing latency by 60% and compute costs by 40% for production workloads. Implemented monitoring dashboards and alerting systems using Prometheus and Grafana ensuring production reliability, uptime, and proactive incident detection. Project 4: Intact Insurance (Scalable Backend & Data-Driven ML Services) Role: Python Developer | Duration: Jan 2018 Jul 2020 Location: Chicago, IL Environment: Python, SQL, PySpark, Pandas, NumPy, scikit-learn, TensorFlow, XGBoost, LightGBM, Docker, FastAPI, Flask Developed scalable backend applications and data-driven services using Python, implementing modular architecture, reusable components, and robust error handling for production environments serving insurance claim processing. Designed and optimized high-volume data pipelines using Python and Apache Spark, processing large datasets while improving runtime performance through efficient memory management and parallel execution strategies. Built RESTful APIs using FastAPI/Flask to expose machine learning models and business logic, enabling real-time data access for downstream applications with comprehensive API documentation. Wrote complex SQL queries and integrated Python with relational databases to support high-performance data retrieval, transformation, and analytics for insurance underwriting models. Developed data processing utilities using Pandas and NumPy for cleansing, transformation, and feature engineering across multiple business datasets including claims, policies, and customer records. Containerized Python applications using Docker and supported CI/CD pipelines through Git-based version control and automated deployments with quality gates and testing stages. Integrated cloud services with Python applications for scalable compute, storage, and event-driven processing supporting insurance analytics workloads. Built logging, monitoring, and alerting mechanisms to proactively detect failures and maintain production reliability across all deployed services. Collaborated with cross-functional teams to translate business requirements into technical solutions, delivering production-grade software within agile environments with sprint planning and retrospectives. Improved application performance by profiling bottlenecks, optimizing algorithms, and implementing asynchronous processing where applicable, achieving 3x throughput improvements on key pipelines. PRIORITY PROJECTS LLM Infrastructure & Hosting Platform Technologies: Hugging Face Transformers, Triton Inference Server, Kubernetes, Vector Databases, PyTorch, TensorRT Hands-on LLM training and fine-tuning implementations using LoRA/QLoRA parameter-efficient techniques for domain-specific model adaptation Production GPU-centric model server design and Kubernetes orchestration with StatefulSets for persistent model state management Custom embedding pipelines and tokenization optimization using SentencePiece for domain adaptation with vocabulary extension Low-latency API inference architecture with sub-second response times using Triton Inference Server with dynamic batching and model ensemble Comprehensive LLM observability dashboards tracking performance metrics, token throughput, GPU utilization, and reliability KPIs ML Model Operationalization -- Production Deployment Technologies: Python, PySpark, scikit-learn, VADER, Pandas, MLflow, Docker, Kubernetes Production regression models and classification pipelines with automated deployment, versioning, and rollback capabilities Advanced NLP models including TF-IDF vectorization, topic modeling with Gensim, and sentiment analysis pipelines Distributed PySpark ML pipelines for large-scale training on Databricks clusters processing terabytes of data Comprehensive Model Risk Oversight dashboards with drift detection using KS tests, PSI calculations, and feature importance monitoring Seamless business integration of ML services with REST APIs, batch processing, and event-driven architectures CERTIFICATIONS AWS Machine Learning -- Specialty -- Amazon Web Services Google Cloud Machine Learning Engineer -- Google Cloud TensorFlow Developer Certificate -- Google PyTorch Developer Certification -- Meta EDUCATION Master s / Bachelor s in Computer Science / Engineering Keywords: continuous integration continuous deployment artificial intelligence machine learning sthree rlang Illinois Kansas