Avani Khemka
Career Objective
Building scalable intelligent systems and solving real-world engineering challenges. I am a final-year Computer Science Engineering (AI) student focused on building production-ready systems, not just prototypes. With a strong foundation in Machine Learning, Data Science, and Embedded Systems, I thrive on solving complex problems—from automating recruitment workflows to optimizing sensor anomaly detection. I value clean architecture, impact-driven engineering, and continuous learning.
Education
B.Tech in Computer Science Engineering (AI)
JECRC Foundation
2022 – 2026
8.94 CGPA
Skills
Trainings & Experience
Smart India Hackathon 2023 & 2024, Team Lead & Lead Engineer
2023 – 2024Led cross-functional teams in two consecutive national hackathons. Architected full-stack solutions for government problem statements (DRDO & Ministry of Tourism). Managed development sprints, code reviews, and final deployment.
PwC Launchpad, Technology Trainee
Mar 2025 - Jun 2025Selected via competitive AI-proctored assessment. Gained deep exposure to Cybersecurity protocols, Enterprise AI integration, and Python automation. Completed specialized micro-certifications.
Upflairs, Embedded Systems Trainee
Aug 2023Developed automated control systems using embedded C and Python. Bridged hardware-software interface for real-time sensor data processing.
Certifications
- Certified Application Developer (CAD) – ServiceNow 2025
- Certified System Administrator (CSA) – ServiceNow 2025
- Embedded Systems Certification – Upflairs 2023
Projects Undertaken
Recruitment Automation System
- Technologies Used: Python, Spacy (NLP), TF-IDF, FastAPI, React.
- Recruiters spend 70% of their time manually screening thousands of resumes, leading to fatigue and bias.
- Built an NLP pipeline to extract resume entities (skills, exp) and calculate contextual semantic similarity against job descriptions. Automated interview scheduling via Google Calendar API.
Sensor Data Anomaly Detection
- Technologies Used: Python, Scikit-Learn, Pandas, Matplotlib.
- Detecting rare equipment failures in noisy sensor data is critical to prevent downtime but difficult due to class imbalance.
- Engineered a hybrid ML approach using Isolation Forests and Autoencoders. Preprocessed high-dimensional time-series data to isolate signal noise.
ghumogaon
- Technologies Used: React, Node.js, Express, MongoDB, Google Maps API.
- Tourists overcrowd popular spots while hidden gems remain economically stagnant due to lack of visibility.
- Designed a recommendation engine that surfaces 'underrated' locations based on user preferences. Implemented secure booking flow and interactive maps.
Job Application Tracker
- Technologies Used: Python, Tkinter, SQLite, Pandas.
- Spreadsheets are tedious and manual for tracking hundreds of job applications.
- Created a local database application with a GUI wrapper. Features include status updates (Applied, Interview, Offer). Features include adding jobs, viewing statuses, and simple management.