Jan 2024 – Apr 2024 · 3 mos · Austin, Texas, United States · Remote

Feb 2023 – Jan 2024 · 11 mos · Austin, Texas, United States · Remote

• Designed core infrastructure for MLOPs platform which involves end-end model development lifecyle.
• Implemented multiple MLOPs lifecycle POCs with ML libraries & Jupyter notebooks using MLFlow.
• Onboarded LLMOPs lifecycle design to the core traditional Machine Learning MLOPs lifecycle.
• Developed CI/CD pipelines for platform components to ensure continuous integration.
• Responsible for library and language version upgrades and fix security vulnerabilities as required.
• Collaborated with product owner to define requirements for MVP feature stories for sprints and releases.
• Improve developer tooling & infrastructure, maintain dev clusters (Kubernetes).

Jan 2022 – Dec 2022 · 11 mos

• Working as a generalist software engineer and acting team lead for Ethical AI model scanning product.
• End-to-end AI application development life cycle as a backend developer for various cloud platforms.
• Developed CI/CD pipelines from scratch to build development and QA environments by improving continuous integration pipelines and deployment over Kubernetes.
• Automated infrastructure, applications, and services to enable an automated delivery through the CI/CD pipelines to the cloud.
• Designed an end-end product release process which was responsible for multiple successful releases.
• Working with the Product Owner to define requirements for MVP and clearly define feature stories for sprints and releases.
• Working with SRE and QA teams to ensure high availability infrastructure and product quality.
• Debugging and resolving customer deployment bugs and helping them install/set up smoothly on the cloud and on-prem infrastructure.
• Tech Stack:
• Python
• Node.js / TypeScript
• Golang
• Kubernetes
• GoCD
• Postgress
• AWS
• GCP
• Azure

Apr 2019 – Jan 2022 · 2 yrs 9 mos

• Currently working as a part of AI platform team. The aim of the platform is to make it easy for Machine Learning Developers, Data Scientists, and Data Engineers to take their ML projects from ideation to production and deployment, quickly and cost-effectively.
• Technology stack:
• Python
• Node.js
• Go-lang
• Kubernetes
• MongoDB
• Redis
• RabbitMQ

May 2018 – Mar 2019 · 10 mos

Apr 2016 – Apr 2018 · 2 yrs

• Responsibilities:
• Improving the data quality using stop-words removal, eliminating syntactic anomalies, decomposing and assembling the data.
• Data deduplication using Cosine Similarity and Lavenshtein distance algorithms to transform the data into a unified schema.
• Implemented Focus Extraction, quantity parser and concept extraction using Stanford’s CoreNLP components (POS Tagger, Dependency parser, NER, Lemmatization and Tokenization) on the user query.
• Implemented Type Model: A module which connects the NLP extracted fields (focus, quantity, concepts) to the aggregated values to find the similar n-grams.
• Designed and implemented end-to-end product search pipeline on Elasticsearch.
• Implemented Lemmatized and tokenized indexing of documents, which makes search cleaner and handles variations of words like tenses, plurals etc.
• Used function score to modify the relevance score of the documents based on the price, portability, size and other factors. This increases the probability of the relevant documents to be ranked higher.
• Technologies Used:
• Languages: Python 3.6, Java
• Frameworks/Tools: Flask + Tornado, Play, Asyncio
• NLP modules: Stanford CoreNLP, Spacy
• Databases: MongoDB, Redis
• Deployment Tools: Docker, Rancher, GoCD
• Architecture: Microservice architecture

Jun 2014 – Apr 2016 · 1 yr 10 mos

• Responsibilities:
• Designing and developing REST APIs in python using Flask framework.
• Optimising flask micro-services to serve large number of concurrent requests.
• Designing distributed pipelines for large scale web scraping using RabbitMQ, Celery, Redis and Scrapy.
• Deploying micro-services as Docker containers.
• Building event-driven micro-services.
• Designing database schemas in NoSQL Databases like MongoDB.
• Graph modelling for optimised search with large number of connections using Neo4j.
• Extracting entities from unstructured text, connecting them via graph using Neo4j, figuring out unseen relations between data.
• Technologies Used:
• Languages: Python
• Frameworks: Scrapy, BeautifulSoup, RabbitMQ, Celery
• Databases: Redis, MongoDB, Neo4J