Canada
I am currently working as Lead AI/LLM Engineer with a strong foundation as a Machine Learning Engineer, bringing 5+ years of professional experience in leading, building, and deploying tailored machine learning solutions across diverse business use cases. My expertise spans designing ML systems, software engineering, MLOps, and deploying ML applications on the cloud. I have successfully contributed to projects for four enterprise clients and have played a key role in developing more than three in-house initiatives. I hold an MSc in Computer Science from Lakehead University, graduating with a GPA of 3.67. Skills • Programming Language: Python, SQL, NOSQL, R, Bash, Shell Scripting • Machine Learning & Deep Learning: TensorFlow, PyTorch, Keras, NumPy, Pandas, Scikit-Learn, OpenCV, NLTK, Spacy • Data Science Tools: ETL Pipeline, Data Science Pipeline (Data Cleaning, Wrangling, Visualization, Modelling, Analysis), Tableau, PySpark • Development Tools: MLOps, GIT, Elasticsearch, ELK Stack, Redis, Kafka, Boto3, Jupyter, Pip/Conda, Dynaconf, Cookie Cutter • Deployment Tools: Docker, Docker Compose, Kubernetes • Cloud Technology: Amazon Web Service (AWS), Google Cloud Platform (GCP) • Others: FastAPI, Flask, Django, JavaScript, Load Testing, Python Design Patterns, Jira
• Leading team of engineers in architecting and developing LLM based solutions for Pinterest.
•Worked on prompt engineering and labelling automation projects from Pinterest.
•Involved in data extraction, data wrangling, data cleaning, analyzing data, prompt development, discovering patterns of misclassified samples, prompt deployment and orchestration. • Deployed prompt to production for different projects that labelled around 20 million+ data samples, estimated saving of ~$3M in labelling cost and more than 5 months in time as compared to human labellers. • Orchestrated labelling workflow using Airflow that consisted of extraction of data from database, triggering prompt labelling and prompt chaining for large-scale data labelling. • Worked on Instruction based Fine-Tuning for multi modal LLM using QLoRA for classifying relevance between Image and Search Query.
Worked on Information Extraction project for large media company to extract, process, analyze and automate user feedback process. Used Topic Modelling, Clustering and LLMs to uncover specific service issues with paper, lowered the manual analysis and correlation with various user data or business decision. Worked on planning, designing and creating learning materials and learning pathways of Software Engineering and MLOps topics for training and upskilling in-house engineers.
Worked on Research and Development of in-house cloud native MlOps platform on-top of available open-source MLOps tools for versioning, serving and monitoring. Developed python packages to serve and control models with KServe in Kubernetes, researched and worked on creating custom serving containers with KServe. Also, created MLOps package for dynamically routing inference request in API. Worked on analyzing the serverless system used for Resume Parsing and analyzing the limitations using load testing tool as well as customizing it for handling the serverless architecture. Also, provided batching solution for the concurrency limitation in the system. Leading Software Engineering Focus Group team who are responsible for designing learning pathways for upskilling and training ML Engineer about SE concepts in ML careers.
Worked on Cyber security project to develop credentials Search API on massive unstructured data. Worked on AWS services like Elasticsearch service, ElasticCache, DocumentDB. Worked on containerization of API. Worked on developing Search API to full text search on various fields of phishing file. Worked on AWS services like API Gateway, lambda function, S3. Worked on NER, Text Extraction and Fine Tuning Bert based model for Text Classification. Also, generated solution for vector (Image Embedding) search with custom analyzers in Elasticsearch, worked on distributed processing of stream of message using Kafka and various search query and analyzer optimization for Elasticsearch.
Supervised the major Final Year Project of undergraduates for the module Project and Professionalism. Supervised and mentored Computer Vision Projects(Face Detection, Face Recognition, Age Estimation), Recommender System Projects(Collaborative filtering, Content-based filtering), Web development