Hi, I’m

Yagna Karthik Vaka

AI Engineer, Sr. Software Engineer
View Resume

About Me

Senior Software Engineer with experience building GenAI- and ML-powered applications, owning modern frontend systems. Delivered LLM-driven, agentic AI features via scalable APIs and React/Next.js interfaces. Strong applied ML background with a Master’s in Data Engineering (Uppsala University), including a research thesis at RISE.

Skills

Core Languages

Python, JavaScript, TypeScript, C++

Machine Learning & AI

Supervised & Unsupervised ML, Deep Learning, PyTorch, Scikit-learn

LLMs & Generative AI

Prompt Engineering, RAG, Gemini, GPT-4o, LangChain, Vector Databases (Chroma)

Reinforcement Learning

Q-Learning, DQN, PPO, Stable-Baselines3

Backend & Systems

Flask, Node.js, REST APIs, Distributed Systems, Data Pipelines

Data Engineering

Apache Spark, Hadoop, Streaming Systems, ETL Pipelines

Cloud & MLOps

AWS, Docker, CI/CD, Model Deployment, Monitoring

Frontend

React, Next.js, HTML, CSS

Expericence

My Expericence

edvenswa Logo
Sr. Software Engineer

Edvenswa Enterprises | Jun 2025- Present

  • Led frontend modernization from legacy Laravel-based workflows to Next.js, improving page load performance by 30%, accelerating developer iteration speed in production systems.
  • Developed and researched GenAI and agentic-AI systems using LangChain, LangGraph, and FastAPI, integrating LLM-driven workflows into scalable services handling concurrent user requests.

Paapaya Logo
AI Lead

Paapaya | May 2024 – Jun 2025

  • Built and deployed GenAI-enabled web applications on AWS, integrating Flask-based APIs with modern frontend components.
  • Designed and implemented prompt-engineering pipelines to automate dynamic content generation, significantly reducing manual content effort.
  • Applied NLP techniques and BERT-based semantic retrieval to improve search relevance and template matching accuracy.

Edvenswa Logo
Sr. Software Engineer

Edvenswa Enterprises | Jan 2022 – Aug 2023

  • Built and maintained backend services and APIs for HIPAA-compliant healthcare platforms, integrating multiple EHR systems for Inferscience using Laravel and Amazon S3.
  • Developed browser-based healthcare tools enabling seamless data exchange between frontend clients and backend systems.
  • Worked with SQL-based data stores and AWS cloud storage to support secure ingestion and retrieval of sensitive healthcare data.
  • Recognized with the “Deep Skill Adder Award” for successful integration of complex EHR workflows.

Edvenswa Logo
Programmer Analyst

Edvenswa Enterprises | Sep 2021 – Jan 2022

  • Developed full-stack, API-driven features for myBenefits.ai using Node.js and TypeScript, enabling scalable backend–frontend integration.
  • Enabled secure, low-latency search and retrieval across 1M+ U.S. physician records using AWS Elasticsearch and backend query optimization.
  • Built mobile and web applications using React Native, integrating backend services for scalable data access in production environments.

Edvenswa Logo
Full Stack Developer

Edvenswa Enterprises | Jan 2021 – Sep 2021

  • Built end-to-end MERN-style applications integrating RESTful APIs, Firestore (NoSQL), and frontend clients for production workflows.
  • Designed backend services supporting real-time data synchronization and user notifications using Firebase Cloud Messaging.
  • Developed and deployed a custom admin backend to manage users, messaging, and core application workflows in production environments.

Education

Uppsala University
Uppsala University

Uppsala, Sweden

Master of Science in Data Engineering

2023 - 2025

Indian Institute of Information Technology Sricity
Indian Institute of Information Technology Sricity

Andhra Pradesh, India

Bachelor of Technology in Computer Science

2017 - 2021

Research & Projects

RISE Logo
Master’s Thesis Student

RISE | Jan 2025 – Jun 2025

  • Designed and evaluated reinforcement learning (RL) strategies to optimize privacy–utility trade-offs in data-driven systems under adversarial settings.
  • Implemented and benchmarked advanced RL algorithms against inference attacks, analyzing performance, training stability, and utility degradation.

UU Logo
From Pixels to Material Properties: ML for Solar Cell Analysis

Uppsala University | Aug 2024 – Jan 2025

  • Developed an end-to-end ML pipeline to detect phase boundaries in solar cell materials for a self-driving laboratory.
  • Built scalable image preprocessing and alignment pipelines with custom DataGenerator for efficient batch loading.
  • Designed and trained a U-Net–based deep learning model with 8-channel inputs and dynamic learning-rate scheduling, achieving 97.17% test accuracy.

UU Logo
Movie Expert Chatbot with Vector Database (RAG)

Uppsala University | Aug 2024 – Oct 2024

  • Built a Retrieval-Augmented Generation (RAG) chatbot using Gemini 1.5 LLM and Chroma vector database for personalized movie recommendations.
  • Implemented semantic search, multimodal querying, and context-aware responses using Python and Streamlit.
  • Containerized the system with Docker, enabling reproducible deployment and persistent chat history.

UU Logo
Predicting GitHub Stargazer Trends

Uppsala University | Mar 2024 – Jun 2024

  • Designed a scalable ML pipeline to predict GitHub repository popularity using historical metadata and activity signals.
  • Achieved R² scores of 0.846 (Random Forest) and 0.831 (Gradient Boosting) through feature engineering and hyperparameter tuning with Ray Tune.
  • Automated model training and deployment using Docker and Ansible within a CI/CD workflow.

UU Logo
Reinforcement Learning for Atari Pong

Uppsala University | Mar 2024 – Jun 2024

  • Implemented Deep Q-Network (DQN) agents with epsilon-greedy and NoisyNet exploration strategies for Atari environments.
  • Trained and optimized agents on GPU using PyTorch, improving performance from −21 to +21 reward.
  • Designed reusable RL architectures enabling rapid adaptation across multiple Atari games.

UU Logo
Bike Availability Prediction (Washington, D.C.)

Uppsala University | Dec 2023 – Jan 2024

  • Analyzed multi-year Capital Bikeshare data to model bike availability patterns across Washington, D.C.
  • Implemented ensemble learning techniques (Bagging), achieving 88.2% prediction accuracy and ranking 4th overall.
  • Built data pipelines and models using Pandas and Scikit-learn.

IIITS Logo
Visual Question Answering (VQA) on VizWiz Dataset

Indian Institute of Information Technology, Sricity | Jan 2020 – Dec 2020

  • Built a multimodal VQA system combining computer vision and NLP to answer questions about real-world images.
  • Implemented and evaluated attention-based architectures (Bottom-Up, MCAN, MUTAN), achieving up to 55.08% accuracy.
  • Trained deep learning models using PyTorch on the VizWiz dataset with GPU acceleration.

IIITS Logo
Speech Recognition System

Indian Institute of Information Technology, Sricity | Aug 2018 – Dec 2018

  • Built a speech recognition system using MFCC feature extraction and classical ML models.
  • Trained and evaluated models in MATLAB with robustness analysis under noisy conditions.
  • Awarded best project and top grade for technical quality and performance.

Achievements & Certifications

Coordinated as a Host for Utnarm in the Uppsala University

Utnarm
View Credential

R language for Data Science With Real Exercises.

Udemy
View Credential

Python for Data Science & ML Bootcamp.

Udemy
View Credential

Integrating Multiple EHRs for HealthCare Domain
Deep Skill Adder Reward

Edvenswa Enterprises
View Credential

Deep Learning NLP: Training GPT-2 from Scratch

Coursera
View Credential

Twitter Sentiment Analysis using NLP

Coursera
View Credential

Ranked among Top 400 participants in National Level Online Hackathon

Bridgingo
View Credential

Reach Out to me!

Let's Connect and discuss !!!

Hyderabad, India