Building intelligent, production-ready AI systems

Deekshitha
Kotte

AI/ML Engineer with experience designing data pipelines, LLM-powered automation, and cloud-native solutions across enterprise environments.

AI / ML & Generative AI LLMs, RAG, & automation Python · Cloud · MLOps
View LinkedIn Profile ↗ ✉ Get in touch
Machine Learning LLMs & RAG Python Cloud & Data Pipelines MLOps
About

Who I am as an engineer

I work across data, models, and engineering to deliver AI solutions that are practical, measurable, and grounded in real-world constraints.

I specialize in AI and ML engineering, with experience building data pipelines, model-driven services, and integration layers that bring intelligence into existing products and workflows.

My background spans model development, experimentation, and the engineering needed to make those models useful in production: clean APIs, monitoring, and feedback loops.

I’m particularly interested in using LLMs and retrieval-augmented systems to automate repetitive analysis, accelerate decision-making, and improve how teams interact with their data.

  • Hands-on experience across the ML lifecycle: data prep, training, evaluation, and deployment.
  • Comfortable collaborating with product, data, and engineering teams to scope and deliver solutions.
  • Strong focus on clarity, reliability, and writing code that is easy for others to understand and extend.
Skills

What I work with

A mix of AI/ML, data, and engineering tools used to build real applications.

AI / ML

Developing and integrating ML models into real products.

Machine Learning Model Training Evaluation Experimentation

LLMs & RAG

Using large language models over structured and unstructured data.

LLMs RAG Prompting AI Assistants

Programming

Writing clean, maintainable code for data and backend systems.

Python APIs Automation Scripting

Data & Pipelines

Preparing, transforming, and serving data for AI workloads.

Data Pipelines ETL Analytics SQL / NoSQL

Cloud & Deployment

Running models and services in cloud environments.

Cloud Platforms Containers CI/CD

Collaboration

Working across teams to define, build, and iterate on AI solutions.

Cross-functional Documentation Code Reviews
Projects

Things I’ve built

Real-world machine learning and computer vision projects showcasing deep learning, model training, and applied AI problem-solving.

Clinical AI + RAG Pipelines

Healthcare-Grade Clinical Document Intelligence (RAG + LLM)

Summarizes clinical notes, flags safety risks, and provides grounded differential diagnoses.

Built an end-to-end Retrieval-Augmented Generation (RAG) system for clinical document analysis. The assistant ingests de-identified clinical notes, builds a FAISS vector index, and generates evidence-backed summaries, safety risk signals, and differential diagnoses using LLMs grounded in retrieved medical context. Includes a FastAPI backend and a React-based clinician workflow UI.

FastAPI Python LangChain OpenAI FAISS React Docker
View on GitHub ↗
Computer Vision + Multi-Modal Models

AI-Powered Multi-Modal Cancer Detection

Combines image + structured data for tumor classification.

Designed a deep-learning system integrating ResNet50 for MRI image analysis with an MLP network for structured patient features. Built the full pipeline — ingestion, preprocessing, training, and unified inference — to evaluate multi-modal cancer detection performance.

Python TensorFlow/Keras ResNet50 MLP Deep Learning
View on GitHub ↗
Medical Imaging · CNNs · AI Diagnostics

MRI Tumor Detector

Transfer learning model that detects tumors from MRI scans.

Developed a tumor-detection pipeline using preprocessing, augmentation, and fine-tuned CNN architecture for MRI classification. Focused on model accuracy, clean data workflows, and reproducible experimentation.

Python TensorFlow/Keras Transfer Learning Medical Imaging
View on GitHub ↗
Experience

Where I’ve been learning and building

Roles focused on AI, ML, and software to support real products and teams.

AI / ML Engineer
Various Projects & Roles
4+ years
  • Designed and deployed RAG pipelines using Azure OpenAI, LangChain, and FAISS for document summarization and intelligent automation.
  • Collaborated with teams to understand requirements and translate them into model and system designs.
  • Improved clinical workflow automation by 60% through AI-driven context retrieval and generative summarization pipelines.
  • Designed multi-modal AI models (CNN + Transformer) for document understanding and entity extraction.
  • Built modular LLM services integrated with backend APIs using FastAPI and deployed on AKS with CI/CD automation.
  • Helped evaluate and iterate on model performance using metrics, experiments, and feedback.
  • Supported integration of ML components into applications, focusing on reliability and usability.
Software & Data Engineering
Engineering & Analytics Work
Previous roles
  • Developed high-performance REST APIs and event-driven microservices for AI data ingestion using .NET Core and Kafka.
  • Integrated front-end and back-end systems through RESTful APIs, improving data exchange efficiency by 25%.
  • Built and maintained scripts, services, and tools that supported analytics and reporting.
  • Worked with data pipelines to ensure that information used by models and dashboards was accurate.
  • Automated end-to-end CI/CD pipelines with Azure DevOps improving release frequency by 35%.
  • Participated in code reviews, documentation, and knowledge sharing within engineering teams.
  • Helped debug and improve existing systems with a focus on clarity and maintainability.
Contact

Let’s connect

Open to AI/ML and software roles where I can contribute to building thoughtful, useful, and reliable systems.

LinkedIn
linkedin.com/in/deekshitha-k
Email
kdeekshitha92@gmail.com
Links
LinkedIn GitHub

Feel free to reach out if you’d like to discuss roles, projects, or collaboration opportunities.