Senior AI Engineer — Nuremberg
I'm Pranay, I design agent systems, orchestration, and the platforms that carry them from prototype into production. After hours, I shoot cinematic travel film and play sports.
01
6+
Years shipping production AI
02
3
Patents in industrial AI
03
1st
AI hackathon, Fraunhofer IOSB-INA
04
2+ yrs
Mentoring DS & ML, TechLabs
6+ years building production AI,
end-to-end.
From novel research to enterprise platforms.
Wherever models meet reality.
At DB Systel I build agentic-RAG for Deutsche Bahn's GenAI platform. Before that, at Tvarit GmbH I co-authored 3 patents in industrial AI that shipped into real manufacturing plants.
Lifelong Learner.
Experience
Architecture lead for Deutsche Bahn's enterprise RAG platform, owning product strategy, agentic-RAG delivery, and cross-team technical alignment.
Built DB's API-first RAG-as-a-Service platform on Azure with retrieval, multi-tenant capability, and the API gateway layer.
Owned the MLOps backbone including CI/CD, Kubernetes, and ML pipelines.
Invented a patented time-series domain-adaptation algorithm, supervised a team of three junior researchers.
US Patent · US12032929B2 · Granted Jul. 2024
System and Method for Cross Domain Generalization for Industrial Artificial Intelligence Applications
Adversarial encoder-decoder architecture enabling predictive models to transfer across machines at a fraction of the data and compute cost. Enables quality prediction on new machines without retraining from scratch.
EP Patent · EP4254087B1 · Granted Mar. 2026
System and Method for Recommending a Recipe in a Manufacturing Process
Physics-AI hybrid system: physics-based simulation generates synthetic training data; ML model optimizes manufacturing recipes under real environmental and machine-state variability.
EP Patent · EP4300229A2 · Application Pending
System and Method for Determining Defect Regions of Products in a Manufacturing Process
Geometry-aware ML: integrates 3D CAD mesh features with process sensor data to predict spatial defect locations in manufactured parts.
Shipped predictive models into live manufacturing plants — 50% scrap reduction for a wheel manufacturer.
Deep learning research on multivariate sensor data, novel regularization against catastrophic forgetting.
Projects
Workload & readiness tracking for a T20 cricket squad with ACWR, bowling load, and RPE, with Telegram check-in reminders.
A browser-based constraint-aware school timetable generator.
An LLM-based extraction and classification pipeline for transactions straight from raw bank PDFs/CSVs feeding a dashboard.
Education
Education
Fachhochschule Südwestfalen · Soest, Germany
Machine Learning · Deep Learning · Signal Processing · Advanced Control Systems
Jawaharlal Nehru Technological University · Hyderabad, India
Volunteer Experience
Skills & Stack
01Agentic & GenAI
02Languages & Frameworks
03Cloud & Infra
04MLOps & Data
05Domains
Large Language Models · Time-Series Forecasting · Computer Vision · Deep Learning · Domain Adaptation
After Hours
Same obsessive eye for craft but different medium. I shoot travel film at 24fps, chasing the blue hour from Nuremberg to wherever the light is interesting.
@pranay.studios on InstagramPlaying club cricket in Germany. Batting middle order, keeping wickets, bowling medium pace. The sport that followed me across continents.
Weekly court sessions chasing drop shots and smashes. Fast-twitch reflexes, strategic patience. What I started to play for my wife.
Contact
Open to senior AI engineering roles, consulting, and interesting collaborations. Drop me a line, I get back within a day.
[email protected]