top of page

SATISH BHAMBRI

Master's Software Engineering ASU

  1. Senior Machine Learning Engineer, Walmart Labs                                       (December 2022 - Present)

    • Designed and Developed Conversational GroceryBot Agent using Langchain, GCP and VertexAI Matching Engine for ANN vector similarity search, Implemented grounded and Retreival Augmented Generation LLM(text-bison).

    • Developed NLP models for Smart Ads Recommendation using Topic modelling and transformers like KNN, BERTopic,

      Topic2Vec and Two Towers resulting in a 20% increase in click-through rates.

    • Developed Evaluation metrics for smart ads without customer feedback ground truth using cosine similarity, KeyBERT MMR, Longformers. (Supervised proxy problem and Smoke Tests for unsupervised learning) achieving a

      15% improvement in personalized recommendations.

    • Led experiments to enhance Diversity metrics for recommendations, increasing the diversity score by 25%.

    • Optimized recommendation generation using CuML, RAPIDS, and NVIDIA GPUs, reducing processing time by 30%.

    • Developed Recipe Recommendations with KNN, Content Based and Collaborative filtering, LLMs and Generative AI goes well with section.
       

  2. Senior Data Scientist, Product Development, R&D, BlueYonder.                      (April 2019 - Dec 2022)

    • Developed Risk As A Service, identifying the supply chain disruption hotspots for Natural Disasters and Impacted ports and freight using Azure functions, Java, NLP, Naïve Bayes, Random Forests and LightGBM, leading to potential $2 million in cost savings.

    • Engineered data pipelines for predicting shipment time of arrival, reducing prediction errors by 40%, using microservices, Kafka, Data Lake, Synapse Db, Apache Ignite Cache, and ML models of NLP and Tree based ensemble methods, Bayesian Heuristics.

    • Dockerized Java applications and orchestrated deployments on Azure using DevTest Labs and MULE, resulting in a 20% reduction in deployment time.

    • Developed Automated Integration Estimate Recommendor streamlining customer pitch preparation and increasing customer engagement by 14%.
       

  3. Software Engineer (MuleSoft), Apisero, Chandler, AZ                             (November 2018 – April 2019)

    • Designed and developed Integration Solutions using Mule 4.1 flows, delivering 8 successful integrations for Minneapolis Airport and JDA projects.
       

  4. Software Engineer, Veras Retail, Phoenix, AZ                                             (July 2018 – November 2018)

    • Innovated Deferred Manager Override and Tender-Based Promotions Module, resulting in a 18% increase in customer engagement.

    • Created APIs for Alexa skills, leading to a 11% boost in flash sales revenue and 22% faster EOD status report generation.
       

  5. Software Engineer Intern (Python), Integrated Device Technology            (January 2018 - May 2018)

    • Developed production build system for ARM chip simulation and common clock framework drivers reducing build times by 48%.
       

  6. Software Engineer (Python, .Net), Sapient Global Markets                              (January 2015 - July 2016)

    • Designed and Developed Electricity Trading Platform employing Microservices using WCF, UDDI for Canadian State of Alberta, resulting in a 25% increase in trading efficiency.

    • Managed Version Control using TFS and implemented continuous deployment using Jenkins, reducing deployment failures by 30%.
       

  7. Research and Achievements:
     

    • SIGAME – Computational Astrophysics (Python, Open Source)

    •  Developed module for the prediction of L[CII] – SFR and [Cu] Luminosity Function at the Epoch of Reionization based on the simulation data.
       

    • Intelligent Feedback Analysis System (NLP, Word2Vec, TF-IDF, LDA, Spectral Clustering) 2017- 2018 • A Java Web Application for microprocessor simulation and a feedback analysis system being implemented in Microprocessor lectures providing the current student performance and standing in course using Natural Language Processing.
       

    • Intelligent Team FormaMon System for Researchers (NLP, StaMsMcal Machine Learning, Doc2Vec, KNN) Fall 2017 • Developed a Recommendation Engine using NLP techniques of Doc2Vec and clustering models of Hierarchical, Spectral Clustering and Minimum Sum graph based algorithm for the best match for the given problem.
       

    • School of Earth and Space Exploration (CUDA, Python, Matlab, Astropy) Fall 2016 • Developed GPU-optimized version of a Radio Bursts telescope algorithm using CUDA (from Nvidia), and Dark Matter simulation of galaxies.
      ​

    • Quantum Clouds (Publication)

    • Addressing the problems of cloud via quantum computational model (http://arxiv.org/abs/1410.6502) 

bottom of page