The Data Grill
Welcome to The Data Grill.
The Data Grill is a blog where I write about data platforms and AI as I learn them and put them to work with customers in real-world scenarios. It spans the full data estate, from getting your data ready for agentic AI to the governance that keeps it trustworthy.
This is my space to turn what I learn and build into clear field notes and courses, the kind that help readers understand a concept and apply it to their own work, not just read about it.
Every article and lesson here is written by me, Amrutha Satishkumar, after thorough research and drawn directly from my experience working with Microsoft customers.
Stay tuned, more courses and labs are coming soon.
All posts reflect my personal views and do not represent my employer.
About the author
Hi, I'm Amrutha Satishkumar.
I'm a Data & AI Solution Engineer at Microsoft, based in San Jose, California, working at the intersection of data and AI. I build the platforms that make AI possible and the agentic systems that run on top of them.
Over 5+ years I've grown from deep learning and NLP research into production data engineering and enterprise AI. Along the way I've shipped pipelines that cut training time, models that sharpened forecasting, and AI agents that removed manual work outright, always with an eye on what actually holds up once it leaves the lab.
Today I partner with State Local Government and education customers to modernize their data estates and adopt AI responsibly. Much of that work happens in the room: architecture deep-dives, technical presentations, and live demos that turn ideas into something teams can build on. I care most about solutions that survive past the demo: secure, scalable, governed, and genuinely useful to the people who depend on them.
5+
Years across data & AI
100+
Enterprise customers
35+
Solutions shipped to production
6
Microsoft certifications
AI-Ready Data Foundations
Modern data warehouse architecture, estate modernization, and migration strategy that make enterprise data trustworthy enough for AI.
Agentic AI in Production
Designing and deploying agentic and generative AI systems that move beyond pilots to deliver measurable business outcomes.
Governance & Responsible AI
Security, lineage, access control, and evaluation engineered in from day one, not bolted on later.
From Pilot to Production
Performance, cost optimization, and maintainability so solutions scale and survive long past the demo.
What I care about
Customer obsessed
I start from the outcome the customer needs and work backward, so the solution fits their reality, not just the slide.
Curiosity over comfort
I learn in the open, document what I figure out, and chase the why behind every system I touch.
Substance over hype
I care less about buzzwords and more about what actually ships, scales, and holds up in production.
Trust by design
Governance, security, and clarity are part of the build, never an afterthought bolted on later.
Career
Where I've worked
-
Data & AI Solution Engineer · Microsoft
Dec 2025 to Present
San Jose, California
Trusted technical advisor to state, local, education, and ISV customers across the data and AI lifecycle. I design secure, scalable Azure analytics solutions, lead modernization and migration planning, and translate customer requirements into production-ready architectures.
- Guide adoption of Microsoft Fabric, Databricks, Synapse, Power BI, and Purview
- Lead technical presentations, live demos, and hands-on sessions for customers and universities
- Focus on governance, performance, cost optimization, and long-term maintainability
- Partner with sales, engineering, and product from envisioning to implementation readiness
Microsoft FabricDatabricksSynapsePower BIPurviewAzure -
Associate Data Consultant, Data Platforms & AI · Lantern Microsoft Partner
May 2022 to Nov 2025
Remote
Delivered end-to-end data and AI solutions on Microsoft Azure at Lantern, a Microsoft partner, spanning agentic AI, NLP, and applied machine learning from data pipelines through production deployment and Power BI dashboards.
- Built across the Microsoft Azure data and AI stack as a Microsoft partner, from Azure Machine Learning and Azure OpenAI to Microsoft Fabric and Azure DevOps
- Agentic AI document automation cut manual review time by 70%
- Azure DevOps CI/CD reduced release errors by 40%
- NLP litigation risk modeling reduced legal costs by 20%
Azure MLAzure OpenAIMicrosoft FabricPower BIMLOpsAzure DevOps -
Data Science Intern · Lantern Microsoft Partner
Jan 2022 to Apr 2022
Dallas, Texas
Embedded with the data science team to build and validate ML models end to end, from feature engineering through evaluation and handoff to production pipelines.
- Engineered features and ran model selection across classification and regression tasks for client analytics use cases
- Built and cleaned training datasets to improve signal quality and reduce downstream model noise
- Collaborated directly with senior data scientists, accelerating the path from experimentation to deployment
PythonPandasScikit-learnSQL -
Data Analyst, Project · TXU Energy
Sep 2022 to Dec 2022
Dallas, Texas
Delivered analytics for one of Texas's largest energy providers, translating raw subscription and customer data into executive-ready dashboards that drove commercial strategy.
- Designed Tableau dashboards tracking KPIs for residential and commercial electricity-plan subscriptions
- Built and automated Alteryx data prep workflows to replace manual reporting and accelerate insight delivery
- Surfaced trends informing pricing and product decisions for business stakeholders
TableauAlteryxSQL -
Data Analyst Intern · Hewlett Packard Enterprise
Jan 2021 to Jul 2021
Remote
Worked within HPE's global analytics function to optimize BI infrastructure and improve the accuracy of ML models powering operational reporting for leadership.
- Redesigned ETL pipelines, accelerating data delivery by 15% across reporting workflows
- Improved Qlik app reload performance by 20% through query and data model optimization
- Applied SVM feature engineering to boost predictive model performance by 20%
- Built dashboards enabling global leadership to track operational KPIs in real time
QlikTableauPythonSQLETL -
Data Science Developer · Center for Pattern Recognition & Machine Intelligence
Jan 2019 to Dec 2020
Bengaluru, India
Researched and built deep learning and NLP systems across medical imaging and literary text analysis, delivering models that reached production-grade accuracy benchmarks.
- Built a COVID-19 detection model using VGG16 and PyTorch on X-ray and EHR data, achieving 98% accuracy
- Developed NLP pipelines for persona and relationship extraction using NER and sentiment analysis, lifting reader engagement by 25%
- Designed full training pipelines from data ingestion through evaluation and results reporting
PyTorchPythonNLPDeep LearningVGG16 -
Cloud Developer Intern · Center for Cloud Computing & Big Data, PES University
Jun 2018 to Dec 2018
Bangalore, India
Researched distributed resource allocation and built real-time big data ingestion prototypes.
- Prototyped a decentralized, partition-based resource framework with Docker and Mesos APIs
- Designed a Spark Streaming and Hadoop model for live, parallel data processing
SparkHadoopDockerMesosKubernetes
Capabilities
What I work with
Azure Data Platforms
AI & Agents
Analytics & BI
Governance & Delivery
Certifications
-
Azure Data Scientist Associate
Microsoft Certified
-
Fabric Analytics Engineer Associate
Microsoft Certified
-
Azure Data Engineer Associate
Microsoft Certified
-
Azure AI Fundamentals
Microsoft Certified
-
Azure Data Fundamentals
Microsoft Certified
-
Azure Fundamentals
Microsoft Certified
Education
-
MS in Business Analytics, Specialization in Data Platforms & AI
Southern Methodist University, Cox School of Business
2021 to 2022
-
BTech in Computer Science Engineering, Minor in Data Science
PES University
2017 to 2021