Hi, I'm Anshuman Dey Kirty
Data Engineer | AI Engineer | Software Engineer: Data Infrastructure
I design scalable data platforms and AI-native systems from streaming pipelines and semantic layers to agents that ship to production. Specializing in all modern data technologies with 5+ years of experience.
About Me
My Mission: Track macro AI trends through data‑driven systems and modeling, empowering organizations to stay ahead and build for tomorrow.
I've spent five years making data behave where the stakes are real: Million $ oil rigs, banks, pharma and other Fortune 500 companies. Lately that work keeps converging on one problem: an AI agent is only as trustworthy as the layer beneath it. So I build that governed semantic layer, data models and lead teams doing the same.
Interests
Skills
Resume
Professional Summary
Experienced professional with a strong background in designing and delivering scalable solutions for complex business challenges by leveraging data, including governed AI agents deployed in production at scale. Proven track record of optimizing operations, automating processes, and supporting impactful outcomes for major clients across multiple industries through data-driven approaches. Recognized for driving efficiency, reducing manual work, and contributing to high-value projects in dynamic environments using data as the foundation for decision-making and innovation.
Work Experience
Enable Fortune 500 companies to build for tomorrow by building data systems across industries.
- Led the data platform for a production agentic AI assistant at a Top 5 pharma, powering pre‑call planning and post‑call intelligence for ~1,000 field users; designed the semantic layer for agent context engineering and the infrastructure for secure territory‑level access.
- Streamed 1‑second drilling sensor data into a ML feature store, cutting Non‑Productive Time and protecting $45 M in annual oil‑rig ops.
- Deployed active‑learning tagging for insurance catalogs, auto‑labeling 100 % concept matches and slashing manual review 80 %.
- Automated Snowflake + DBT financial reporting, shrinking a 5‑day close to a 15‑minute pipeline run.
- Built a Spark pipeline over 500 M supply‑chain records, driving ML‑based part‑to‑vehicle prioritization for a global automaker.
Architected ML‑powered log monitoring pipelines, anomaly detection models, AI‑as‑a‑Service microservices, and automated KPI reporting.
- Cut critical incidents 30% YoY by deploying ML log‑monitoring pipelines and pre‑empting 20+ outages with Isolation‑Forest anomaly detection.
- Co‑created an "AI‑as‑a‑Service" microservice with the CoE, putting data‑driven decisions at every team's fingertips.
- Automated KPI reports on risk, performance, obsolescence, and vulnerabilities, saving executives 10+ hours weekly.
Education
Master of Science in Computer Science (Data Science)
University of Southern California (USC), Los Angeles 2021 - 2023Graduated Magna Cum Laude, focused on Data Science and Software Engineering
Blog
Get In Touch
Let's Connect
I'm always interested in hearing about new opportunities and interesting projects. Whether you want to collaborate or just say hello, feel free to reach out!