Experience
Research Intern
Corelight
Built a self-supervised multi-agent system (Claude on Amazon Bedrock) that reverse-engineers and statically analyzes malware through a custom 23-tool MCP server (Ghidra, radare2, capa, FLOSS, androguard) to auto-generate per-family YARA rules, with guardrails isolating the agent from the raw malware. By iteratively analyzing agent transcripts and revising the system's prompting and tool orchestration, I advanced its capabilities to raise detection coverage from 40-60% to 80-98% and cut cost from ~$10 to ~$4 per family run, across a ~38K-sample corpus (138 malware families, 6 file types).
Graduate Researcher
Texas A&M University, Dr. Marcus Botacin's Lab
Researching streaming algorithms for malware analysis, designing and evaluating novel and hybrid detection methods that make YARA rule generation faster and more efficient.
Student Research Assistant
Texas A&M University, Dept. of Soil & Crop Sciences
Collaborating with the GAIA research group on digital agricultural challenges, scaling transfer learning models to analyze multi-spectral soil data on High-Performance Computing (HPRC) clusters for precision agriculture.
Software Development Engineer
Zscaler, Hyderabad, India
Built and enhanced a custom Pytest-based monitoring system running 900+ daily regression tests. Integrated Terraform and Ansible automation to cut environment setup time from 18 hours to 3, and conducted security and maintainability reviews for 70+ pull requests.
Associate Software Developer Engineer
Zscaler, Hyderabad, India
Migrated legacy modules into a unified, modern framework (reducing technical debt by ~50%) and integrated Zscaler Private Access (ZPA) secure-access features adopted by 7+ internal application teams.
Software Developer Intern
Zscaler, Bangalore, India
Automated the lifecycle of simulation environments on AWS (EC2, S3, Secrets Manager), 12x faster, and extended monitoring with Kibana and OpenSearch across 1,000+ system metrics.
Software Developer Intern
Athen Tech India, Hyderabad, India
Built a customer-facing chatbot in JavaScript with state-machine conversation flows and backend API integration, resolving 60%+ of common support queries without human intervention.