Jul 2023 – Aug 2024 · 1 yr 1 mo

• Primary liaison for M.Tech (Research) students in the Placement Committee at IIT Mandi.
• Connected with numerous recruiters to organize campus placement drives involving 200+ companies.
• Facilitated student-recruiter communication, coordinated pre-placement talks, and managed placement-related activities.

Aug 2022 – Present · 3 yrs 10 mos

• Characterized the latent spaces of convolution-based autoencoder variants and demonstrated that the latent representations of clean images learned by CAE, DAE and VAE models lie on a product manifold of symmetric positive semidefinite (SPSD), with CAE and DAE forming stratified product manifolds, arising from variability in the ranks of the SPSD components, whereas VAE exhibiting a smooth product manifold, characterized by a uniform rank structure across samples.
• Examined how these structures behave under clean, noisy, and blurred inputs, with implications for each model’s suitability for smooth interpolation, denoising, and deblurring.
• Analyzed the structure of representations learned by MAE through self-supervised pretraining and subsequent supervised fine-tuning, with the goal of identifying geometric properties that support their strong classification performance.

Aug 2022 – Present · 3 yrs 10 mos

• Supported following courses at IIT Mandi :-
• EE534 - Probability and Random Processes
• DS401 - Optimization for Data Science (twice)
• DS403 - Introduction to Statistical Learning
• CS202 - Data Structure and Algorithms
• CS207 - Applied Databases Practicum
• Assisted in grading assignments, conducting tutorials, and providing academic support to students.
• Collaborated with professors to develop course materials and facilitate student learning.

Jan 2022 – Jun 2022 · 5 mos · Indore, Madhya Pradesh, India · On-site

• Engineered a BERT-based NLU module for chatbots leveraging Hugging Face Transformers and scikit-learn for named entity recognition and contextual intent analysis, enabling accurate interpretation of user queries in real-time conversations.
• Achieved 83% intent classification accuracy on an internally curated test set by fine-tuning BERT and input preprocessing.

Jan 2021 – Jun 2021 · 5 mos · Bengaluru, Karnataka, India

• Worked on the project Auto-TM PROM Data Verification and Error Detection System in the System Engineering - Controls and Communication Group, U.R. RAO Satellite Centre, ISRO, Bangalore.
• Developed an automated pipeline for the verification process of satellite telemetry (TM) PROM data and generation of structured test error reports, minimizing manpower and manual error by 30% across 5 ground stations using supervised ML.
• Led a 4-person team and implemented an anomaly detection pipeline leveraging the Random Forest algorithm trained on labeled telemetry logs (manually tagged in the past) to flag inconsistent or faulty data transmissions, achieving over 95% accuracy in detection.

May 2019 – Jun 2019 · 1 mo · Bhilai, Chhattisgarh, India · On-site

• Delved into the workings of various departments such as Electric Repair Shop, Electronic Technical Lab, and Telecom at BSP.
• Gained valuable insights into the connectivity of different sectors within a fully integrated industry.