Close

Ajay Kesarwani

Full-Stack AI Engineer · RAG & LangChain · Python

5+ years in telecom backend · MSc Computer Science · Master’s in AI (ongoing)

Download Resume

About Me

I’m a Full-Stack AI Engineer with 5+ years of experience building large-scale backend systems and end‑to‑end AI applications. I specialize in combining robust Java/Python backends with modern LLM technologies, especially RAG (Retrieval-Augmented Generation) and LangChain, to ship production-ready, intelligent features.

Previously, I worked as a Lead Software Engineer at HCL Technologies (client: Cisco) and as a Software Developer at Ericsson, building and maintaining high-availability telecom systems for 4G/5G networks. This experience taught me to design reliable services, debug complex distributed systems, and deliver zero‑downtime releases for critical clients.

I hold an M.Sc. in Computer Science from the University of Passau and am currently pursuing a Master’s in Artificial Intelligence, focusing on ML, NLP, and LLM-based systems. My projects span deep learning, NLP, and RAG/LangChain‑based applications. Recently, I’ve been focusing on building RAG-powered applications with LangChain and modern LLMs.

What I’m looking for

Currently looking for Full‑Stack AI Engineer / AI Engineer roles focusing on RAG, LLMs, and data‑intensive backend systems in Europe (onsite or remote).

Experience

HCL Technologies

Lead Software Engineer

  • Led development of 5G solutions providing a platform for communication between multiple applications, improving performance and reliability for critical telecom services.
  • Upgraded and refactored core modules, managed a small development team, and delivered releases with zero downtime for production systems.
  • Acted as the primary point of contact for live application maintenance, diagnosing and resolving high‑priority issues under time pressure.
  • Collaborated with data and AI teams to design services that can support analytics and AI-driven features on top of telecom data.

Ericsson

Software Developer

  • Developed software packages for centralized mobile network provisioning, supporting large‑scale 4G/5G deployments and live application troubleshooting.
  • Worked closely with clients to gather requirements, translate them into technical tasks, and coordinate deliverables with the development team.
  • Improved operational workflows for network provisioning and enhancement by implementing robust, maintainable backend components.

Education

University of Passau

April 2025 - Present

Master's in Artificial Intelligence Engineering

University of Passau

Oct 2020 - Sept 2024

Master's in Computer Science

Indian Institute of Engineering Science and Technology (IIEST), Shibpur

July 2010 - June 2014

Bachelor of Engineering in Computer Science

Projects

Thesis project screenshot

Master’s Thesis: Exploring Constructiveness in Online News Comments

Leveraged NLP techniques to analyze comment patterns in online news, achieving a 94% F1-score.

Technology: Python, TensorFlow, Keras, NLTK, spaCy, Pandas, Streamlit

View Project
Passenger Volume Prediction

Passenger Volume Prediction for Bus Travel

Applied time-series forecasting techniques to achieve 82% accuracy in predicting passenger volume

Technology: Python, Scikit-Learn, Tensorflow, AutoARIMA, LSTM

View Project
Accident Classification

Accident Video Classification

Preprocessed video data into image frames; LSTM model outperformed others for accident classification, achieving 82% accuracy.

Technology: Python, OpenCV, TensorFlow, LSTM, CNN, MobileNetV2

View Project
Music Genre Classification

Music Genre Classification

Achieved 78% accuracy with a CNN model using Mel-Frequency Cepstral Coefficients (MFCCs) features using GTZAN dataset.

Technology: Python, Librosa, TensorFlow, CNN

View Project
Web Page Scrapping

Web Scraping and Twitter Follower Analysis

Implemented Python scripts to extract and analyze data from various platforms, including Twitter follower networks. Technology: Python, BeautifulSoup, Selenium, Pandas, Numpy, Matplotlib, Seaborn, JavaScript

View Project
Cold EMail Generator

Cold EMail Generator

Developed an automation tool using LangChain for personalized email generation and sentiment analysis.

Technology: Python, Chromadb, Groq, LangChain, Streamlit, LLM

View Project
Restaurant Name Generator

Restaurant Name Generator

Built a tool to generate restaurant names and menus, utilizing LLMs and a temperature slider.

Technology: Python, Groq, LangChain, Streamlit, LLM

View Project

Skills

Get in Touch