Machine Learning Engineer

Gowtham Senthilnayaki

Boston, MA (Open to Relocate) 857-***-**** **@**.*** LinkedIn Portfolio GitHub SUMMARY

Machine Learning Engineer with 3+ years of experience building and deploying end-to-end ML systems using LLMs, GANs, and transformer architectures across text, image, and multimodal domains. Proficient in PyTorch, TensorFlow, FastAPI, and scalable model deployment with MLOps tools across cloud environments. Delivered production-ready solutions for recommendation, document understanding, generative QA, and smart interaction systems. EDUCATION

Boston University Sep 2022 - Jan 2024

Master of Science in Artificial Intelligence Boston, MA Sikkim Manipal University July 2016 - Jun 2020

Bachelor of Technology in Computer Science and Engineering India SKILLS

Skills: Forecasting, RAG, Causal Inference, Experimentation Techniques, Agentic LLM Tools, Data Visualization. Languages: Python, MATLAB, C, C++, Swift, R, Julia, Java, Scala, SQL, PostgreSQL. Frameworks/Libraries: TensorFlow, Keras, PyTorch, NodeJS, Django, SciPy, Scikit-learn, JAX, GAN, Airflow, FastAPI, LangChain, TogetherAI, LamaIndex, Agentic AI, Ray, Terraform. Databases/Tools: Docker, MongoDB, Spark, Tableau, AWS, Google Cloud Platform (GCP), Azure, CrewAI, MLFlow, Git. Testing: PyTest, JUnit, Selenium, Postman, Unit test, Integration Test, A/B Test. EXPERIENCE

Data Scientist III (ML) - DermaSkan May 2024 - Present

• Developed an MVP and a data pipeline processing gigabytes of data every 3 days using pandas and numpy, reducing data processing time by 40% and improving operational efficiency.

• Implemented and monitored collaborative filtering and neural network-based recommendation models with Keras, achieving a 25% improvement in product match accuracy as validated through A/B testing. Machine Learning/AI Engineer - Data Inception(Consultant)Mar 2024 - May 2024

• Built a CNN-based OCR engine with PyTorch and FastAPI, enhancing data processing accuracy by over 15%.

• Implemented retrieval-augmented generation with LLAMA for conversational QA tasks, enhancing document-grounded response generation accuracy by 25%.

Research Assistant (NLP / AI) - Boston University Jun 2023 - Jan 2024

• Fine-tuned a large language model using BERT and AWS SageMaker, achieving a 20% improvement in insights extraction as measured by a ROUGE score of 0.6.

• Built graph embeddings with Node2Vec in Python to enhance consumer spending analytics, enabling better-targeted marketing strategies.

Lead Machine Learning/AI Engineer - Stealth Mar 2021 - Aug 2022

• Designed a resume extractor using BERT embeddings and AWS, improving keyword matching accuracy to 80% and reducing resume screening time by 40%.

• Led a team of 4 to implement a collaborative filtering recommendation engine and ELT data pipeline in Azure, boosting user-event matching by 30% and increasing engagement by 20%. Data Scientist - Mastech Infotrellis Jun 2020 - Feb 2021

• Deployed a CNN-based OCR pipeline on AWS SageMaker with Docker, achieving an 88% data extraction accuracy and reducing error rates for B2B client insurance claims data processing and leading a client demo of the POC.

• Developed a knowledge graph-based decision support system using graph traversal algorithms, improving querying efficiency by 10% and reducing decision-making time.

AI Intern Research - Bobble.ai Jan 2020 - Jun 2020

• Enhanced speech recognition accuracy for Indian languages from 71% to 82% using a wav2letter model and processing over 1 million speech data points weekly.

• Innovated a CycleGAN-based artistic sticker generator, increasing user engagement through improved visual appeal. PROJECTS

Large Language Model Based Projects - MS AI Mini Project - Code Aug 2023

• Designed and developed a Large Language model that predicts paper selection chances in publications.

• Implemented feedback mechanisms leveraging the LLAMA model and Retrieval-Augmented Generation (RAG) to fine-tune the model, improving prediction accuracy by 15%.

Real Time Image to Image Translation - MS AI Mini Project - Code Apr 2023

• Researched and implemented a more efficient GAN-based Image to Image translation model, coordinating with a team to apply different Computer Vision architectures like Cycle and DCGAN on Facades Dataset. PUBLICATIONS

• T.K. Dugar, Gowtham. S, U.K. Chakraborty. Hyperparameter Tuning using enhanced Authorship Identification using Deep Neural Networks (paper id: ICOEI 227, DOI: 10.1109/ICOEI.2019.8862631) in proceedings of the 3rd International Conference on Trends in Electronics and Informatics.

Contact this candidate