Senior ML Engineer at Eurowings Digital GmbH

I architect enterprise AI systems at scale

Dr. Santhoshkumar · Senior ML Engineer

Architecting enterprise AI systems: search, RAG, and MLOps platforms serving millions with 40ms P99 latency

I architect enterprise AI systems at scale. I reduce time-to-production from months to days through reusable ML infrastructure, turn research insights into enterprise systems, and own end-to-end delivery from requirements through operations.

Search & RetrievalLLM SystemsProduction MLOpsDatabricksAzure
View Resume

Eurowings Digital GmbHEurope | Remote

Dr. Santhoshkumar

Production AI Leadership

Senior ML Engineer / Data Scientist

Current Focus

Search, LLMs, and MLOps

Production
13+
Years

2012-2016: Backend (Java) Engineering → Team Lead
2016-2020: PhD (ML) + Team Leadership
2020-2026: Enterprise AI Systems at scale in Lufthansa Group

4
Featured Production Systems

Recent flagship projects: Search, RAG, MLOps, Forecasting. 10K QPS, 2M embeddings, 40ms P99 latency.

Proven Impact

  • 18% support call reduction = $420K annual savings
  • 45% search satisfaction improvement (62% → 90%)
  • 98% deployment time reduction (weeks → hours)

Featured Projects

Production AI systems built for enterprise scale. From search to RAG to MLOps infrastructure.

⭐ Flagship Project
Eurowings · Lufthansa Group4-month production rollout

AI-Powered Semantic Search

Enterprise search powered by embeddings, retrieval, and LLMs

Why it matters

Keyword-only search caused poor retrieval for natural-language questions, multilingual queries, and support intent inside booking journeys.

Built an onsite search and FAQ discovery platform for airline e-commerce, handling natural-language search, multilingual queries, and hybrid retrieval at customer scale.

Architecture Highlights

  • Nuxt.js website and backend search API
  • Databricks query understanding for language detection, spelling correction, and intent extraction
  • Hybrid BM25 + vector retrieval with confidence scoring and reranking
Semantic SearchRAGVector SearchGPT-4LlamaIndexDatabricks

Impact

  • 18% fewer support calls
  • 45% higher satisfaction
  • 40% lower latency
Eurowings · Lufthansa Group3-month production rollout

Enterprise AI Customer Support

Conversational AI for enterprise customer support

Designed and deployed a customer-support chatbot over enterprise help content, reducing repetitive questions while maintaining citation-backed responses.

Architecture Highlights

  • Azure OpenAI GPT-4 with retrieval-augmented answer generation
  • Hybrid search over FAQs, policy documents, and support content
  • Session-aware conversation memory for follow-up questions
RAGGPT-4LlamaIndexConversational AIAzure

Impact

  • Reduced repetitive tickets
  • Faster resolution
  • Higher self-service rate

Core Challenge

Customers repeatedly asked about baggage, cancellations, refunds, check-in, and special assistance, overloading support channels.

Eurowings · Lufthansa Group6-month platform build

Enterprise MLOps Platform

Infrastructure for operating ML at scale

Built the platform layer for governed ML deployment on Azure Databricks, standardizing CI/CD, registry workflows, monitoring, and access control across teams.

Architecture Highlights

  • Git and Azure DevOps for versioned pipelines
  • Databricks MLflow tracking and model registry
  • Unity Catalog and Feature Store for governance
MLOpsDatabricksMLflowUnity CatalogAzure DevOps

Impact

  • Weeks → Hours deployment
  • Standardized deployments
  • Improved governance

Core Challenge

Teams deployed models inconsistently, with weak governance, fragmented monitoring, and no reproducible delivery path.

Eurowings · Lufthansa Group5-month production rollout

AI Forecasting & Personalization

Predictive modeling for business decisions

Applied forecasting and recommendation pipelines to improve planning accuracy, targeting quality, and campaign efficiency across digital operations.

Architecture Highlights

  • Feature pipelines for reusable forecasting inputs
  • Ensemble forecasting with XGBoost, Prophet, Random Forest, and regression models
  • Collaborative and similarity-based recommendation components for personalization
XGBoostProphetCollaborative FilteringTime Series

Impact

  • Improved forecast accuracy
  • Better targeting
  • Reduced manual effort

Core Challenge

Manual forecasting and non-personalized campaigns created planning inefficiencies and weaker customer targeting.

Impact at Scale

Production systems, measurable outcomes, and engineering ownership across search, LLM applications, and platform delivery.

13+
Years Experience
4
Production AI Systems
3
LLM Applications
Enterprise
MLOps Projects

Leadership and Credibility

Senior-level signals from career progression, doctoral depth, and enterprise ownership.

PhD in Machine Learning

VIT, Vellore · 2020 · GPA 9/10

Completed doctoral research on rumor detection and control with deep learning.

4 Years Team Leadership

Team Lead, AEM Development · 2016-2020

Led software delivery teams while completing PhD research in parallel.

13+ Years in Industry

From Java/AEM to Production AI

Progressed from enterprise software engineering to ownership of production AI systems.

Selected Production Initiatives

Four production systems that show how I work across retrieval, LLM applications, platform engineering, and applied machine learning.

4-month production rollout

AI-Powered Semantic Search

Eurowings · Lufthansa Group
  • Keyword-only search caused poor retrieval for natural-language questions, multilingual queries, and support intent inside booking journeys.
  • 18% reduction in customer support calls
  • 45% increase in search satisfaction
Semantic SearchRAGVector SearchGPT-4LlamaIndex
3-month production rollout

Enterprise AI Customer Support

Eurowings · Lufthansa Group
  • Customers repeatedly asked about baggage, cancellations, refunds, check-in, and special assistance, overloading support channels.
  • Reduced repetitive support workload
  • Improved first-response speed
RAGGPT-4LlamaIndexConversational AIAzure
6-month platform build

Enterprise MLOps Platform

Eurowings · Lufthansa Group
  • Teams deployed models inconsistently, with weak governance, fragmented monitoring, and no reproducible delivery path.
  • Reduced deployment time from weeks to hours
  • Standardized delivery across ML teams
MLOpsDatabricksMLflowUnity CatalogAzure DevOps
5-month production rollout

AI Forecasting & Personalization

Eurowings · Lufthansa Group
  • Manual forecasting and non-personalized campaigns created planning inefficiencies and weaker customer targeting.
  • Improved forecast accuracy for planning workflows
  • Raised campaign targeting quality
XGBoostProphetCollaborative FilteringTime Series

Speaking and Credentials

Teaching, invited talks, and certifications that reinforce technical leadership.

International Faculty Development Program (6-days)

Gopalan College of Engineering and Management (GCEM)2025-02-03 to 2025-02-08Bengaluru, India

Delivered NLP and data science sessions for faculty, with research directions and live demos.

Guest Lecture on Professional Software Development Using Java

School of Information Technology and Engineering2020-02-05VIT, Vellore

Presented software engineering process, tools, and applied Java practices.

Conference Proceedings talk on Rumor Control

INDIA-20182018-07-20Mauritius

Presented pulse-vaccination based rumor control research.

Faculty development program poster oneFaculty development program poster two

Tech Stack

Capabilities organized by system responsibility rather than a flat keyword list.

Generative AI & LLMs

GPT-4Azure OpenAILlamaIndexLangChainRAGPrompt EngineeringVector SearchSemantic Search

ML Engineering

PythonPyTorchTensorFlowScikit-learnXGBoostMLflowModel Serving

MLOps & Platform

DatabricksAzure MLDockerKubernetesCI/CDModel MonitoringUnity Catalog

Data & Search

Vector DatabasesAzure AI SearchElasticsearchSQLNoSQLData Pipelines

Cloud & Infrastructure

AzureAWSFastAPIREST APIsMicroservices

Selected Publications

Optional reading for deeper technical context and academic work.

2019

International Journal on Semantic Web and Information Systems (IJSWIS)

A Parallel Neural Network approach for Faster Rumor Identification in Online Social Networks

CNN architecture for early rumor identification.

View publication
2020

Journal of Organizational and End User Computing

A Bio-inspired Defensive Rumor Confinement Strategy in Online Social Networks

Bio-inspired strategy for rumor containment.

View publication
2020

International Journal of Web Services Research (IJWSR)

A Neuro-Fuzzy approach to detect Rumors in Online Social Networks

Hybrid neuro-fuzzy model for rumor detection.

View publication
2019

SN Applied Sciences (Springer)

Interest Aware Influential Information Disseminators in Social Networks

Influence modeling for targeted information diffusion.

View publication
2019

Information Systems Design and Intelligent Applications (Springer)

An Effective Rumor Control Approach for Online Social Networks

Pulse-vaccination inspired rumor control strategy.

View publication

Let's Connect

Interested in search, LLM systems, production ML, or enterprise AI architecture? I am open to technical conversations, advisory work, and senior engineering opportunities.

santhoshramuk@gmail.comEurope | Remote