← Back to Home

Experience

Nursing & Engineering Lab

Research Software Engineer

Nursing & Engineering LabAmherst, MA

Sep 2025 - Dec 2025

  • Identified and corrected logical and numerical errors in an embedded EMG signal processing pipeline
  • Fixed incorrect variance and RMS computations by repairing buffer indexing and statistical logic
  • Improved runtime performance by ~50% through buffer management and control-flow optimizations, enabling reliable embedded deployment
CEmbedded SystemsSignal Processing
SciQuel

Software Engineer Intern

SciQuelRemote

May 2025 - Sep 2025

  • Engineered RESTful APIs with Typescript and Next.js / React.js for posting, fetching, and handling data on scientific article contributors from a MongoDB database and rendering them automatically on article webpages.
  • Re-designed and implemented database schemas through Prisma to separate scientific contributors from website users while maintaining the usabilities of existing APIs, enhancing backend efficiency.
Next.jsTypeScriptPrismaTailwind
Forschungszentrum Jülich

Research Assistant

Forschungszentrum JülichJülich, Germany

May 2025 - Aug 2025

  • Worked with Jim Buffat on deep learning & remote sensing for sun-induced fluorescence retrieval at the Forschungszentrum Jülich, iAS-8 (Data Analytics and Machine Learning group)
  • Supported by the DAAD RISE Scholarship.
PyTorchPython
Build UMass

Software Engineer

Build UMassAmherst, MA

Feb 2025 - May 2025

  • Built and deployed an AI-powered chatbot for NEFAC to automate user support and streamline information access.
  • Optimized backend architecture using FastAPI and GraphQL, reducing API response times by 40% and scaling to handle 20+ concurrent queries/sec; integrated with a React frontend for intuitive user experiences.
  • Automated memory and context retrieval using FAISS and LangChain, powering real-time agentic query resolution.
TypeScriptFAISSLangChainFastAPI
Salesforce

AI / ML Fellow

SalesforceAmherst, MA

Aug 2024 - Dec 2024

  • Developed AI models for SAAS customer sentiment analysis and lead conversion prediction with 91% accuracy using Random Forest.
  • Utilized Prophet for time series forecasting to optimize outreach timing for predictive analytics.
PythonPandasProphetSQL
Tufts University

Research Assistant

Tufts UniversityMedford, MA

May 2024 - Aug 2024

  • Worked with Prof. James Murphy on problems related to hyperspectral image clustering & unmixing.
  • Supported by the Tufts VERSE REU.
PyTorchPythonMATLAB