Nachiket Kathoke

Data Scientist

Kharagpur, West Bengal, India2 yrs 10 mos experience

Most Likely To SwitchAI ML Practitioner

Key Highlights

Expertise in AI applications across audio, images, and text.
Strong fundamentals in CNN, Transformers, and Deep Learning.
Proficient in Python and C++ for coding.

Stackforce AI infers this person is a Data Scientist with expertise in AI and Machine Learning applications.

Contact

Skills

Core Skills

NlpMachine LearningDeep LearningImage Processing

Other Skills

3D U-NetAIAttention MechanismsAudio-lighting SynchronizationClassificationCommunicationConvolutional Neural Networks (CNN)DLData AugmentationData augmentationDependency ParsingDetectionDigital Signal ProcessingEngineeringGraph Transformer

About

I am a final year undergraduate student pursuing B.Tech major in Aerospace engineering and micro specialization in AI & application. Experienced in LLM-powered and AI applications across Audio, Images, and textual domains. Strong fundamentals of CNN, Transformers, and Deep Neural-Nets and skilled with the use of Pytorch, Tensorflow, and open-source tools. I use Python and C++ to jot down code.

Experience

2 yrs 10 mos

Total Experience

1 yr 5 mos

Average Tenure

1 yr 10 mos

Current Experience

Swiggy

Data Scientist-I

Aug 2024 – Present · 1 yr 10 mos · Bengaluru, Karnataka, India · Hybrid

Indian institute of technology, kharagpur

Bachelor Thesis Project

Aug 2023 – Aug 2024 · 1 yr · Kharagpur, West Bengal, India

BTP II :
Deployed a dependency parsing model to analyze the morphological structure and word relationships and
dependency among them in sentences for 3 native languages Tamil, Telugu, and Sanskrit.
Evaluated baseline model performance, enhanced performance by training the RNGTr Graph Transformer model, utilizing attention mechanisms, resulting in a +12% increase in Attachment Score compared to the baseline
Devised a Greedy algorithm to extract the Maximum Spanning Tree from predicted dependency graphs
BTP I :
Employing Attention-Based Trans-LIST Linguistically Informed Tokenizer model for Sanskrit Word Segmentation
Deploying pre-processed textual data of segmented Sandhis to train LSTM variants for Name Entity Recognition

Dependency ParsingGraph TransformerAttention MechanismsGreedy AlgorithmLinguistically Informed TokenizerLSTM+2

Myelin foundry

Data Science Intern

May 2023 – Jul 2023 · 2 mos · Bengaluru, Karnataka, India · On-site

Worked on music generation model based on user input vocals, integrating auto-differentiable digital signal processing (DDSP) priors with the deep learning models
Engineered real-time singing vocal (user's) grading algorithm with respect to original vocal based on the pitch estimation using the Deep Learning model
Designed the Audio-lighting synchronization algorithm analyzing different audio features like energy, Spectograms, spectral centroid, MFCC, etc.

Music GenerationDigital Signal ProcessingDeep LearningPitch EstimationAudio-lighting SynchronizationMachine Learning

Basal analytics

AI Developer Intern

Mar 2023 – Apr 2023 · 1 mo · Bengaluru, Karnataka, India

Worked on various tools and technologies, including Python, AI, ML, DL, and
image processing.
Implemented and trained model for semantic segmentation, detection,
classification, and image super-resolution.
Extensively worked on algorithms like U-Net, U2-Net, ESR-GANs, and Res-Net.

PythonAIMLDLImage ProcessingSemantic Segmentation+4

Indian institute of technology, kharagpur

Research Intern

Oct 2022 – Feb 2023 · 4 mos · Kharagpur, West Bengal, India · On-site

Aimed to develop a DL model to generate cranial implant for skull damage under Grand Challenge Auto Implant
Data augmentation by generating artificial defects and automatic co-registration, to train a more robust model
Developed 3D U-Net and Shape Attentive-UNet model to predict implant and checked its accuracy against the DSC and Hausdorff distance

Deep LearningData Augmentation3D U-NetShape Attentive-UNet