work history

Experience

Teaching Assistant — Computer Vision
University of Maryland College Park · CMNS
College Park, MD, USA · Sep 2025 – Present
  • Instructing MSML640 Computer Vision — guiding 60+ graduate students through theoretical and hands-on coursework in image processing, deep learning architectures, and visual recognition pipelines.
  • Deliver timely, in-depth feedback on assignments and quizzes, upholding rigorous grading standards across varied assessment formats and ensuring academic integrity.
  • Facilitate weekly office hours focused on debugging student model implementations and demystifying complex topics such as CNNs, object detection, and spatial feature extraction.
  • Partner with faculty to refine lab exercises and coordinate review sessions, yielding measurable improvements in student performance and long-term concept retention.
Software Engineer Intern
CBTS (Cincinnati Bell Technology Solutions)
Cincinnati, OH, USA · Jun 2025 – Aug 2025
  • Automated vulnerability detection by integrating CVEDetails with AWS Lambda, EventBridge, and PostgreSQL to parse vendor CVEs every 3 hours — eliminating duplicate alerts and ensuring timely Office 365 email notifications.
  • Deployed a GitHub App-based secret scanner using Lambda, SQS, and Fargate to securely clone repos, execute Gitleaks scans, and block PR merges, reducing credential exposure by 60% across enterprise codebases.
  • Implemented an automated notification workflow with Office 365, instantly alerting developers of detected secrets while enforcing branch protection rules — improving security compliance by 80% and cutting manual review effort significantly.
PythonAWS LambdaSQSFargateEventBridgePostgreSQLGitleaksOffice 365
Teaching Assistant — Algorithms & Data Structures for ML
University of Maryland College Park · CMNS
College Park, MD, USA · Jan 2025 – May 2025
  • Supported MSML606 Algorithms & Data Structures for ML — helping 60+ graduate students navigate the mathematical and algorithmic foundations underlying modern machine learning systems.
  • Evaluated student work through comprehensive assignment and quiz grading, providing detailed written explanations that reinforced conceptual understanding and improved subsequent performance.
  • Held weekly office hours to address challenging algorithmic topics including dynamic programming, graph algorithms, and their direct applications in ML model design.
  • Collaborated with faculty to identify recurring student pain-points and adapt session materials, contributing to curriculum refinements that improved course flow and assessment clarity.
Software Engineer
Accenture
Pune, Maharashtra, India · Jul 2022 – Jun 2024
  • Developed 20+ RESTful APIs with Python, Flask, and Swagger, supporting seamless integration for 200+ enterprise clients and enhancing maintainability through rigorous Postman-based testing.
  • Automated GCP deployments with Terraform + Docker, resolved 100+ Artifact Registry issues, and decreased deployment time by ~40% while stabilizing build and publish pipelines.
  • Designed 20+ real-time entitlement dashboards in Looker with LookML, providing stakeholders with live visibility into client access metrics for SIEM/SOAR workflows.
  • Identified and fixed 40+ critical API bugs; authored 80+ PyTest unit test cases to ensure end-to-end reliability and stability of all created APIs.
  • Built Python solutions for BigQuery integrations, Cloud Scheduler jobs, and 10+ Cloud Monitoring alert policies to proactively monitor SIEM and SOAR metrics.
  • Authored 50+ POC documents to accelerate cross-team knowledge transfer; enhanced sprint efficiency by 20% through Agile ceremonies, 80+ user stories written, and 70+ completed.
Python/FlaskDockerTerraformBigQueryLooker/LookMLPostgreSQLPyTestGitHub Actions
Research Analyst — Propulsion Engineering Centre
Cranfield University
Cranfield, UK · Jan 2022 – May 2022
  • Developed an engine performance data analytics system using MATLAB and Turbomatch, simulating diverse flight conditions to achieve 95% accuracy in predicting key engine parameters for digital twin maintenance decisions.
  • Implemented scattered interpolation algorithms to process remote engine telemetry, cutting measurement error rates to <1% (Case 1) and ~3% (Case 2), ensuring reliable performance assessments across conditions.
MATLABTurbomatchData InterpolationDigital Twin