Hey there! Thank you for checking out my repositories. Here is a little section to highlight some of my projects and learnings, along with a little something about my professional experience.
- Languages: Python, SAP ABAP, Android
- Cloud computing (AWS, SageMaker)
- Frameworks: HuggingFace, PyTorch
- Database: MySQL, RedShift, DuckDB
- Vector Store: ChromaDB
- Natural Language Processing
| Lead Data Scientist | Aug. 2021 – Dec. 2023 |
|---|
- Led a team of four data scientists in the development of various classification models.
- Designed and executed proof-of-concept for the analysis of social media text using transformer-based models.
- Created and maintained custom tables to store model details including the data that the models were trained on. Built code to find and deploy the latest (validated) model for inferencing.
- Created templates for the team to use for the implementation of models. The template was created using best practices of programming and configurable parameters.
- Helped the Data Engineering team design input data tables and logic for the extraction of relevant data. Worked with the Visualization team in creating tables and views that would be used on their dashboards.
- Used the Agile methodology with the development team thereby helping with the launch of new versions and updates.
- Tools like JIRA were used to track and report the development progress.
- Created templates in Confluence for documentation of the models. Documentation was maintained and checked by peers to ensure that all the latest details (version number, version status, etc.) were up-to-date.
- Built serverless solutions to ingest data using AWS SageMaker and pipeline programs.
- Built a solution for the classification of data using EC2 machines. Amazon EventBridge and Lambda functions were used to achieve a solution that could be used by functional teams to execute jobs without worrying about system costs.
- Worked in Microsoft Azure to create a PoC for the classification of product types using Chat-GPT. Evaluated time, accuracy, and costs between version 4 and 3.5-turbo.
- Exploring solutions for the generation of data using Open Large Language Models like Llama.
- Exploring trade-offs between LLMs using agents and tools in CrewAI. Connected three agents with a DuckDuckGo API to retrieve data from the internet.
- Implemented a solution using custom embedding functions and a vector store called ChromaDB for the storage of key phrases and querying the store for similar sentences/key phrases.
| Software Engineer and Director: Scattered Dots (Proprietorship) | Sep. 2014 – Aug. 2021 |
|---|
- Designed and developed an application on Android for the collection of market research data. Custom questionnaires could be programmed and configured to be downloaded as a study. The data collected is then transmitted to a hosted MySQL server.
- Develop new products for text data analysis using Transformers and models like BERT and GPT-3.
- Review quality (via extensive EDA using Python libraries such as Matplotlib) and relevance of collected data for business use cases before committing them to a model for training, review, and updates.
- Conduct data analysis using supervised and unsupervised Machine Learning (statistical) techniques such as regression (simple, multiple, logistic), classification, and clustering.
- Developed a serverless API (FastAPI) to serve a zero-shot text classifier using Docker and Google Cloud.
- Developed business intelligence reports in PowerBI to present learnings from the NLP services.
- Proficient in Jupyter Notebooks and Google Colaboratory to develop prototypes and proof of concepts - as well as PyCharm for production-level development.
- Designed computer vision models for logo classification (Convolutional Neural Networks), achieving 97% accuracy and a recall score of 90%, correctly predicting the logos on which it was trained.
- Directly accountable for the development, leadership, and progression of a dynamic team of software engineers and functional and marketing experts; engage with clients, and develop clear specifications for well-organized and strategic product plans.
- Support teams in creating critical proof of concept, and rapidly prototyping new data capture capabilities.
| Senior Systems Analyst | Sep. 2011 – Aug. 2014 |
|---|
- Travelled to client location as an expert in SAP ABAP, gathering requirements for the technical team, and developments that required a quick turnaround time.
- Led a team of 5 developers for the migration of data to a new version of SAP. Completed this project with a success rate of 98.4% of the data moved in the first session. Analyzed the errors in real-time and created reports which helped in the correction of logic and movement of the remaining data.
- Led teams of developers (different technical areas) in a maintenance project. Assigned and tracked the work to my developers and reported progress to the management.
- Developed plans for a Brown-fields project including the list of objects that need to be developed. Phase assignment and timelines with Gantt chart.
| Systems Analyst | Aug. 2006 – Aug. 2011 |
|---|
- Worked at a Center of Excellence, compiling objects by using industry best practices in creating a library of reusable objects. Projects using this library reduced their overall deployment time by up to 54% when using this library
- Created Reports, Forms, and Enhancements for Finance, Materials Management, and Sales & Distribution.
- Trained developers in Greenfield programs on LSMWs, SAP-Script, and Smartforms. Average trainer rating of 4.6 out of 5.
- Recognized in 2021 and 2022 for my contribution and efforts for the PoC and expertise brought to the project.
- Awarded for my contribution to project success by leading my team to a 98.4% success rate in migrating data into their globally implemented SAP system.
- Mentored a team that created a level-one autonomous driving program. This paper was selected and presented at the "International Conference on Secure Cyber Computing and Communications" in 2021
- Finished in the top three in the "Battlesnake Virtual Coding Event (Canada)" in 2020. https://play.battlesnake.com/events/communitech/communitech-rookie/brackets/
- Mentored a project on identification of inconsistent interaction between two individuals, which was presented at the "International Conference on Signal and Data Processing" in 2019
- My project on Logo Comparison was selected and presented at "Cypher" in 2018
| Degree | University | Date |
|---|---|---|
| Post Graduate Certificate in Big Data and Machine Learning | Great Learning | 2017-2018 |
| Masters in Computer Application | Bangalore University | 2003-2006 |
| Bachelor of Science, Physics, Mathematics, and Computer Science | Bangalore University | 2000-2003 |