PK
PRAKASH • MLOps
Production MLOps • GenAI • Azure

Designing ML systems that behave like reliable infrastructure.

MLOps Specialist and AI/ML Engineer building reproducible pipelines, LLM systems, and cloud-native architectures with measurable impact across latency, reliability, and cost.

Current
MLOps Specialist · Futurense Technologies
Experience
7+ years · AI/ML, MLOps, Aviation
Contact for senior roles View GitHub
Prakash Kantumutchu - MLOps Engineer
Prakash Kantumutchu
MLOps Specialist · AI/ML Architect
Azure ML · AKS · ADF MLflow · Airflow · DVC LLMs · RAG · LangChain
● Production-grade systems, not demos.
Profile

From aircraft to AI systems

Background in aircraft maintenance and engineering, now designing robust MLOps and GenAI systems with the same discipline for reliability, safety, and traceability.

I build and operate production-grade AI, ML, and MLOps systems end-to-end—from data ingestion and experimentation to deployment, monitoring, and cost optimization.

Currently at Futurense Technologies, enabling teams to ship reliable ML pipelines on Azure with MLflow, Airflow, Kubernetes, and CI/CD. Previously at Boven Technologies, delivering ML and LLM solutions across classification, prediction, and NLP workloads.

Before AI/ML, I maintained Airbus A320 aircraft at Air India, which shaped how I think about safety, observability, and operational excellence—principles I now apply to ML infrastructure.

Current role
MLOps
Futurense Technologies · India
Experience
7+ yrs
AI/ML, MLOps, aviation engineering
Education
M.Tech
Cryogenics & Vacuum · NIT Rourkela
Location
India
Jalandhar, Punjab (open to remote/hybrid)
Experience

Owning ML systems end-to-end

Roles focused on building reproducible pipelines, GenAI capabilities, and cloud-native architectures with strong reliability and governance requirements.

MLOps Specialist
Futurense Technologies · Noida, India
May 2025 – Present
Azure · MLflow · AKS
  • Built reproducible pipelines with Airflow, MLflow, Docker, and Kubernetes, reducing onboarding time for engineers and teams.
  • Designed cloud-native ML architectures on Azure integrating Cognitive Services, Azure ML, Functions, API Management, and secure VNet deployments.
  • Led LLM and GenAI capability-building using Hugging Face, LangChain, and Azure OpenAI across teams.
  • Developed end-to-end deployment frameworks with CI/CD and automated validation for model releases.
AI/ML Engineer
Boven Technologies · Bengaluru, India
Sep 2023 – May 2025
TensorFlow · PyTorch · PySpark
  • Delivered ML systems using TensorFlow, PyTorch, scikit-learn, and PySpark across classification, prediction, and NLP pipelines.
  • Built and fine-tuned LLMs for summarization, Q&A, and internal automation, reducing manual workload for clients.
  • Engineered MLflow-based lifecycle management with GitHub Actions and Docker for consistent builds and rollback safety.
  • Created FastAPI and Streamlit applications for real-time model access and stakeholder demos.
  • Implemented SHAP and LIME analysis pipelines to meet responsible AI requirements.
Aircraft Maintenance Engineer
Air India Limited · Kolkata, India
Feb 2019 – May 2023
Airbus A320 · CFM56-5B
  • Performed scheduled and unscheduled maintenance on Airbus A320 aircraft, specializing in CFM56-5B engines, APUs, and hydraulic systems.
  • Conducted inspections per DGCA and EASA standards, ensuring compliance and airworthiness.
  • Authored diagnostic reports and coordinated technical documentation audits to improve operational efficiency.
Assistant Professor
Vitam College of Engineering · Visakhapatnam, India
May 2017 – Sep 2018
Teaching · Mentorship
  • Mentored robotics teams, guiding interdisciplinary projects and fostering innovation.
  • Served as Campus Recruitment Officer, coordinating industry outreach and student placements.
Projects

Systems built for production, not demos

Selected projects that demonstrate end-to-end thinking—from data and experimentation to deployment, monitoring, and security in real environments.

Azure Lakehouse: End-to-End Data & AI System

Designed a Medallion-architecture pipeline (Bronze → Silver → Gold) with incremental, metadata-driven ingestion, transformation, and quality enforcement ready for downstream analytics and AI workloads.

Azure Data Factory ADLS Gen2 Azure Databricks Delta Live Tables Unity Catalog Synapse Power BI

→ Automated metadata-driven pipelines delivering ~60% reduction in manual engineering effort.

View details↗

Full MLOps Pipeline: Azure · Airflow · MLflow · AKS

Architected a production-grade MLOps workflow with experiment tracking, data versioning, orchestrated training, containerized deployment, and metrics-driven monitoring across environments.

MLflow Apache Airflow DVC Docker Azure ML AKS Prometheus Grafana

→ Blue-green deployments, automated model promotion, and monitoring for drift, latency, and throughput.

View details↗

Enterprise-Grade RAG System on Azure

Built a Retrieval-Augmented Generation system using Azure OpenAI, embeddings, Azure AI Search, and LangChain, with Redis caching and Azure AD-based access control for enterprise document corpora.

Azure OpenAI Azure AI Search LangChain Vector DB Redis Cache FastAPI

→ Low-latency retrieval with secure role-based access and reranking for higher answer relevance.

View details↗

Model Deployment & Monitoring Framework

Implemented a framework for deploying ML models via containerized APIs with CI/CD, automated tests, and responsible AI components for interpretability and monitoring.

FastAPI Docker GitHub Actions Streamlit SHAP LIME Grafana

→ Added explainability, regression checks, and visibility into performance for every deployed model.

View details↗
Stack

Tools used in real systems

Focused on technologies that ship reliably in production: MLOps tooling, cloud services, LLM frameworks, and observability tooling.

MLOps & Deployment
MLflow Airflow DVC Docker Kubernetes GitHub Actions CI/CD
Cloud & Data
Azure ML Azure Data Factory Azure Databricks ADLS Gen2 Synapse Delta Lake Power BI
Generative AI & LLMs
Azure OpenAI LangChain HuggingFace RAG systems Embeddings Prompt Engineering
Machine Learning
TensorFlow PyTorch scikit-learn PySpark NLP spaCy Transformers
Data & Analytics
Python SQL Pandas NumPy Matplotlib Seaborn
APIs & Monitoring
FastAPI Streamlit REST APIs Prometheus Grafana Logging & Observability
Contact

Let’s talk about ML systems

Open to senior engineering roles focused on MLOps, GenAI, and production ML systems— especially where reliability, scale, and clarity of ownership matter.

Email
k.prakashofficial@gmail.com
Phone
+91-9330980131
GitHub
github.com/prakashkantumutchu
LinkedIn
linkedin.com/in/prakashkantumutchu
Languages
Telugu (first), Hindi, English, Bengali