Shashwat Negi
I'm proud Badger
Who's Building This?
Applied AI Engineer and Data Scientist with a strong Software Engineering foundation, specializing in building production-grade LLM systems, multimodal AI solutions, and scalable cloud infrastructure.
With 3+ years of experience across AI engineering and software development, I deliver measurable business impact through end-to-end ML pipelines, agentic frameworks, and enterprise deployments.
- Developed LangGraph-based agentic systems with GPT-4/Claude routing, achieving 35% cost reduction and 90% automation in enterprise workflows, deployed via FastAPI microservices and AWS Bedrock.
- Engineered production ML models (XGBoost, DBSCAN clustering) with MLflow tracking and A/B testing, delivering 15% conversion lift and building scalable ETL pipelines processing 100K+ daily records on AWS (Airflow, Lambda, EMR).
- AWS Solutions Architect certified with hands-on expertise in cloud-native architectures using Terraform, Kubernetes (EKS), serverless computing, and compliance frameworks (GDPR, HIPAA).
Resume
My journey so far – the experiences I've gained, the skills I've picked up, and the milestones that have shaped who I am today
Education
University of Wisconsin, Madison
Expected May 2026
Master of Science in Data Science; WiSH Scholarship
GPA: 3.8/4.0
Coursework: Statistical Analysis, Data Structures and Algorithms, Machine Learning, Big Data Systems, Deep Learning for CV
Vellore Institute of Technology, India
Jul 2017 - Jul 2021
Bachelor of Technology in Computer Science; magna cum laude
GPA: 4.0/4.0
Coursework: Data Structure and Algorithms, Operating System, Software Engineering, Database Systems, Internet of Things
Technical Skills
Programming Languages
Proficient in multiple programming languages across data science, web development, and systems programming
- Python, Java, R
- JavaScript, TypeScript (React)
- Linux Bash
Frameworks and Libraries
Expertise in AI/ML frameworks, LLM tools, and deep learning libraries
- LangChain, OpenAI, FastAPI, MLFlow
- TensorFlow, PyTorch, scikit-learn
- HuggingFace Transformers
Data Tools and Technologies
Skilled in cloud infrastructure, big data systems, and DevOps tools
- Terraform (IaC), Kafka, Hadoop, BigQuery
- Docker, Azure Databricks, Grafana
- Git, Jupyter
Certifications
- AWS Cloud Practitioner & Solutions Architect
- Cloud Computing Applications, Part 2: Big Data and Applications in the Cloud - University of Illinois Urbana-Champaign
- Python Data Visualization - Rice University
Professional Experience
Infrrd.ai, San Jose, CA
AI Engineer, Summer Intern
Jun 2025 - Aug 2025
- Improved financial entity extraction by developing a LangGraph-based multimodal agentic prompt optimization framework in Python, featuring NLP classification and extraction with dynamic GPT-4/Claude routing on AWS Bedrock.
- Lowered LLM inference costs by 35% via a human-in-the-loop DPO pipeline using Model Context Protocol server exposed through FastAPI microservice, guiding cost-efficient prompt/model selection, with MongoDB feedback storage.
- Processed 1k document per hour, improving the scalable AI system with Pydantic schema, pytest and LangSmith tracing.
- Cut human annotation by 90% by delivering enterprise-grade Java Spring Boot microservices that invoked PyTorch-finetuned LLMs with optimized prompt templates for accurate data extraction, deployed through Jenkins CI/CD.
ZS Associates, New Delhi, India
Software Engineer, Business Technology Solutions
Aug 2021 - May 2024
- Achieved 15% lift in call-to-conversion by engineering DBSCAN clustering and XGBoost classification models for call-plan and territory alignment with MLflow, deployed via API Gateway + Lambda and evaluated through A/B testing.
- Built scalable ETL pipeline ingesting IQVIA data in compliance with GDPR, HIPAA medical standards and cloud security.
- Cut analytics runtime by 50% by building AWS Airflow (MWAA) DAGs and optimizing SQL reports on PostgreSQL.
- Created SLA-focused Tableau dashboards monitoring KPIs, enabling effective storytelling to non-technical stakeholders.
- Reduced PR review time by enforcing coding standards, debugging issues, and tracking Agile Scrum workflows in Jira.
Software Engineer, Intern
Feb 2021 - Jul 2021
- Increased system uptime and netted a 45% reduction in ops costs by migrating legacy workflows to Apache Airflow on Kubernetes (AWS EKS), refactoring Python and Bash scripts, and building Spark jobs on AWS EMR with auto-scaling.
iSCHOOL, University of Wisconsin-Madison
Teaching Assistant
Sept 2024 - Present | LIS 461: Data and Algorithm & LIS 220: Digital Footprints
- Led discussion sections and office hours for 200+ undergraduate students, providing academic support on data science concepts, algorithms, ethical considerations, and digital footprints.
Projects
Projects I have been working on since high school, focusing on my areas of interest such as Data Science, Deep Learning, Machine Learning, and more
- All
- Software Development
- Artificial Intelligence/Machine Learning
- Data Science
Certification and Recommendations
Technical certifications I have attained, along with recommendations from my professors, manager, and supervisor highlighting my interpersonal and leadership skills
Contact
For any business or personal inquiries, please feel free to reach out to me using the given contact methods, or leave a message through the form below.
Address
514 E Washington Ave, Madison, WI 53703
Call
+1 608 316 0789
negi3@wisc.edu