About me
I build the cloud behind connected devices. At energieleser (Nineti GmbH) in Stuttgart, I design and run the AWS infrastructure for an IoT energy-monitoring product with 2,000+ users, after a year and a half shipping its Flutter apps.
Before moving to Germany for my MSc in Informatik at the University of Stuttgart, I interned at Tata Advanced Systems, researched NLP for sustainability reporting with the University of Oldenburg, and won three hackathons, plus a place in the Smart India Hackathon 2024 finals.
Now I’m adding AI to that foundation, the hard way: building neural networks and a small GPT from scratch before moving on to LLM systems. The goal is to be an engineer who can run AI in production and also understand why it works.
What I do
-
Cloud Infrastructure
AWS IoT Core, Lambda, EC2 and DynamoDB, set up to process real-time device data reliably and at a sensible cost.
-
Backend & APIs
REST and real-time APIs, MQTT device communication, and data pipelines from sensor to dashboard.
-
Mobile Apps
Flutter apps for iOS and Android with live dashboards and device control, in production on both app stores.
-
AI & GenAI
Generative AI projects like AI-designed packaging and compliance assistants, now backed by learning deep learning from first principles.
What I’m working towards
-
Now · Neural nets from scratch
Backpropagation by hand, character-level language models and a small GPT, plus the maths behind them.
-
Next · Deep learning
PyTorch, reproducing published results, and an energy forecasting project on public smart-meter data.
-
Then · LLM systems
RAG with measured accuracy, agents with tools, fine-tuning, and serving models on AWS.
-
Goal · AI engineer & researcher
Production engineering with real ML depth, starting with a master’s thesis in an AI research group.
