Lead ML Engineer with 8+ Years of Experience

Nitish Dhakal

Lead Machine Learning Engineer

******.******@*******.*** 904-***-**** Meridian, ID, United States linkedin.com/in/nitishdhakal

SUMMARY

A highly skilled Lead Machine Learning Engineer with 8+ years of hands-on experience in architecting and deploying large-scale AI-driven systems for enterprise-level applications. Adept in advanced Deep Learning techniques, including the development of Large Language Models (LLMs), Generative AI, and NLP solutions. I specialize in creating sophisticated AI/ML pipelines, from data ingestion to model deployment, while implementing MLOps practices to streamline the model lifecycle and ensure scalability, reliability, and performance in production environments. My expertise spans multiple domains, including Reinforcement Learning (RL), Supervised Learning, and Unsupervised Learning, with proficiency in using state-of-the-art AI/ML frameworks such as TensorFlow, PyTorch, Keras, Scikit-learn, XGBoost, and LightGBM. With extensive experience in deploying models on cloud infrastructures using AWS, Azure, and GCP, I am highly skilled in leveraging containerization and orchestration technologies like Docker and Kubernetes to ensure seamless production-grade deployments. I have a deep understanding of model deployment, real-time inference, and AI model optimization, I lead teams to build production-grade AI solutions that drive innovation and deliver value to the business. I am highly skilled in collaborating with cross-functional teams, including product managers, data scientists, engineering teams, and cloud architects, to ensure alignment between AI technologies and business objectives. I have successfully led the integration of MLflow, GitLab CI, and Kubeflow to automate model training, deployment, and monitoring, ensuring the continuous delivery of AI models in production. My passion for Generative AI, machine learning, and cloud-based AI infrastructures allows me to deliver high-performance solutions that push the boundaries of innovation. PROFESSIONAL EXPERIENCE

Lead AI/ML Enginer MLOps Engineer

Micron Technology

01/2021 – Present

•Lead the development and deployment of large-scale LLMs for generative applications using PyTorch, TensorFlow, and Keras, driving key innovations in AI for real-time text generation and conversational AI applications. Worked with stakeholders to align models with business objectives, ensuring the generated solutions meet customer needs.

•Architected and optimized deep learning models for NLP tasks such as text classification, NER, and question answering, using BERT, GPT-3, and custom Transformer-based architectures, improving model efficiency and accuracy for high-impact applications.

•Designed and implemented a Retrieval-Augmented Generation (RAG) pipeline to enhance LLMs, enabling models to dynamically retrieve external data from cloud databases such as BigQuery and AlloyDB to generate accurate, up-to-date responses, in collaboration with the data engineering and cloud teams.

•Deployed MLOps pipelines using Docker and Kubernetes, ensuring scalable, reproducible, and automated deployments of machine learning models across AWS and GCP environments. Collaborated with the production development team to maintain infrastructure and ensure model deployment integrity.

•Developed and maintained FastAPI-based RESTful APIs for real-time AI inference, integrating multiple models into production systems for seamless user interactions. Coordinated with backend teams to ensure API functionality and scalability.

•Spearheaded the integration of MLflow and GitLab CI to automate model training, validation, and deployment pipelines, ensuring continuous model delivery, versioning, and seamless collaboration across engineering teams for model lifecycle management.

•Optimized existing deep learning models by leveraging Scikit-learn, XGBoost, and LightGBM for hybrid machine learning approaches, combining deep learning with traditional machine learning techniques for better performance and faster processing.

•Implemented monitoring and alerting for model performance and data quality using Prometheus, ensuring proactive management of model drift and improving decision-making speed. Collaborated with data scientists and stakeholders to quickly address performance degradation.

•Applied Reinforcement Learning (RL) techniques in real-time AI agents, enabling more autonomous and context-aware decision-making for dynamic environments, in close coordination with cross- functional teams to refine models based on real-world feedback.

•Led the AI strategy for the organization, working closely with business stakeholders and technical teams to align NLP and machine learning solutions with business objectives. Facilitated discussions with production teams to prioritize AI initiatives and integrate new solutions into production environments.

•Conducted regular hyperparameter optimization and model evaluation, ensuring models stay relevant and perform well in evolving use cases. Collaborated with stakeholders to prioritize improvements based on business needs and feedback.

•Partnered with cross-functional teams, including product management and design teams, to integrate AI-driven features into product workflows and deliver innovative user-facing experiences. Data Engineer

Micron Technology

06/2017 – 04/2021

•Engineered ETL pipelines using Apache Spark and Hadoop to process large-scale datasets, transforming raw data into usable formats for machine learning applications. Collaborated with cross- functional teams to define data requirements and ensure the datasets met the needs of downstream AI applications.

•Developed real-time data pipelines with Apache Kafka and Apache Beam, enabling data streaming and integration with machine learning models for fast predictions and actions. Partnered with ML engineers to ensure seamless data flow from ingestion to model inference.

•Architected cloud-based data lakes and data warehouses on GCP and AWS, utilizing BigQuery, Redshift, and Snowflake to store and process petabytes of data. Coordinated with production teams to ensure that data infrastructure could scale with the growth of machine learning models.

•Built data preprocessing pipelines with Pandas and NumPy, ensuring high-quality datasets were available for model training, feature engineering, and selection. Collaborated with data scientists to define preprocessing workflows and ensure data integrity.

•Automated and optimized data workflows for integration with NLP models, ensuring clean and structured data was delivered to AI teams for deep learning tasks. Worked with cross-functional teams to understand data requirements and optimize pipeline efficiency.

•Designed and developed data transformation workflows in Apache Nifi, and integrated these with cloud data platforms for smooth ingestion and processing. Collaborated with cloud engineering teams to ensure smooth data pipeline deployment in production environments.

•Worked closely with data scientists to provide datasets and APIs for model training, employing tools like FastAPI for serving models and Flask for integrating predictions into real-time systems. Coordinated with backend teams for the deployment of model inference systems.

•Created visualizations and data dashboards for the engineering team using Tableau and D3.js, facilitating real-time insights and performance metrics. Collaborated with product and analytics teams to ensure the visualizations aligned with business goals.

•Led the adoption of MLOps workflows, using Terraform and Ansible to automate infrastructure and deployment of machine learning models at scale. Partnered with cross-functional teams to define the best practices for model versioning and deployment.

•Managed cloud-native infrastructure for data engineering workflows, ensuring scalability and performance in AWS and GCP environments. Collaborated with cloud infrastructure teams to optimize resource usage and reduce costs.

•Implemented cloud-native data analytics solutions using S3, Lambda, and Google Pub/Sub, ensuring easy access and fast processing of data for AI models. Worked closely with AI and data teams to improve data processing speeds and efficiency in production systems.

•Collaborated with production development teams to integrate data pipelines into the production environment, ensuring that machine learning models had real-time access to clean, high-quality data. SKILLS

AI/ML Frameworks & Libraries

PyTorch, TensorFlow, Keras, Scikit-learn,

XGBoost, LightGBM, CatBoost, Fast.al, H2Oai,

Caffe, MXNet

Deinforcement Learning (BL)

Q-learning, Molicy Gradient Methods, Actor-Critic

Models, Deen D Norworks (DQN), Proximal Policy

Optimization (PPO)

API Development & Backend

TastAPI, Flask, Django, gRPC, RESTIal APIs,

GraphQL, Microservices

Model Interpretability & Explainability

SHAP, LIME, Integrated Gradients, Explainable AI

(ΧΑΠ

Natural Language Processing (NLP)

BERT, GPT-3, T5, XLNEL ROBERTA, DisulDERT,

Word2Vec, FastText, Spacy, Named Entity

Recognition (NER) Teat Sununacization

Cloud Platforms

AWS (S3, Lambda, Sage Maker, Redshift), Azure

(Antarve MI, Blob Storage), GCP (BigQuery, Pub/

Sub, Vertex Al, Datallow)

Data Science & Analytics

Pandas, NumPy Scipy, Matplotlib, Seaborn,

Tableau, D3. is

Deep Learning

CNNS, RNNS, LSTMs, GRUs, Transformers, Self

Allention, Auention Mechanisms

Data Engineering

Apache Spark, Hadoop, Kafka, Apache Beam, Nifi

- Flink, Snowflake, BigQuery, Hedshift, MongoDB,

Cassandra

Model Deployment

TensorFlow Serving, TorchServe, ModelDB, Flask,

FastAPI, TensorFlow Lite

Generative Al

GANs (Generative Adversarial Networks), Text-to-

Image Generation (BigGAN, DALL/E), CPT-3, VO-

VAL, StyleGAN, BERT

MLOps

MLflow, DVC (Data Version Contral), KuboFlow,

Kubetlow Pipelines, Jenkins, GitLab CI, CircleCI,

Docker, Kubernetes

Containerization & Orchestration

Docker, Kubernetes, OpenShift, GKK, Helm

EDUCATION

Master's Degree, Computer Science

Boise State University

2015 – 2017

Contact this candidate