Download Resume

MAYANK VYAS

Tempe, AZ | +1(602)5171664 | mvyas7@asu.edu | mayankvyas.vercel.app | /mayankv10 | /Mayank-glitch-cpu

Summary

AI/ML Engineer and Software Engineer with hands-on experience delivering LLM-powered agent systems, scalable ML pipelines, and cloud-native applications. Passionate about intelligent automation and experienced in deploying agentic workflows on cloud infrastructure.

Education

Master of Science, Data Science GPA - 3.61/4.00
Arizona State University, Tempe, AZ Expected May 2026
Relevant Coursework: Probability, Statistics, Machine Learning, Natural Language Processing
Bachelors of Technology, Electrical Engineering GPA - 3.66/4.00
Institute of Infrastructure Technology Research and Management (IIT-RAM), GJ, India May 2024
Relevant Coursework: Data Structures, Algorithms, Computer Networks, Machine Learning
Thesis: MaskRoot - Mask R-CNN thesis for root phenotyping; up to 96.5% segmentation accuracy with successful transfer to non-annotated dataset (<10% error vs. experts).

Work Experience

Software Engineer, Research Jan 2025 - Present
Coral Labs https://coral-lab-asu.github.io/ Tempe, Arizona
  • Engineered Information retrieval RAG pipeline using row level chunking with BM25 and pruning algorithms followed by contrastive learning model on 160K NQ tables to get 93% recall at top 10.
  • Worked on ranking optimization for retrieval results using contrastive loss and contextual embeddings.
  • Developed a Universal prompting methodology SEAR for temporal reasoning tasks on 8 varied datasets to get an average of 92.5% HCS score, beating the SoTa methods like CoT, ToT, etc.
Software Engineer, Machine Learning Architecture May 2022 - Jan 2024
Indian Institute of Information Technology Chennai, TamilNadu
  • Built a Cloud IoT infrastructure using Arduino/Raspberry Pi to reduce transmission time to 0.15 ms.
  • Optimized C++ inference kernels to reduce latency by 35%, enabling real-time anomaly detection with AWS.
  • Deployed an IoT + ML pipeline on Raspberry Pi/Arduino using Tiny ML libraries, achieving 99.97% predictive accuracy while reducing transmission costs by 95%.
  • Designed energy-efficient clustering algorithms and encryption that cut sensor energy use by 82.9%.

Publications

Project Experience

JobHunt-AI (Founding Engineer) Present
  • Architected an end-to-end AI automation pipeline using Elasticsearch for high-speed semantic search and a knowledge graph to map complex relationships between job requirements and user skills.
  • Developed an LLM-based agentic workflow that autonomously parses job descriptions and aligns them with candidate resumes, improving job match relevance by 92% and reducing manual search time by 98%.
  • Launched a production Chrome extension that acquired 12 active users and 9 new users across 2 countries (US, India) within the first month, leveraging Google Analytics to monitor user engagement.
Open Source Contributor (INTEL) Jan 2025
  • Implemented a CI/CD pipeline for a real-time monitoring system, streaming metrics to Grafana to improve system observability and reduce fault detection time by 40%.

Additional

HackASU 2025 – First Place Winner (Sponsored by Anthropic)
  • Built an AI-driven gamified platform that transforms educational questions into interactive, story-driven games using FastAPI, Next.js, and adaptive React logic.
  • Awarded First prize among 250+ teams for technical innovation in EdTech gamification.

Skills