Ansh S.

Key Skills

Software

Don't disclose

OpenCV AI Kit (OAK)

Other

Top Subject Matter

Transportation & Logistics (railway maintenance/predictive maintenance)

Computer Vision (maritime surveillance)

Transportation & Logistics (rail track defect detection)

Top Data Types

Text

Image

Top Task Types

Classification

RLHF

Freelancer Overview

Intern – Central Railway Information Systems (CRIS), New Delhi (May 2025 – July 2025). Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose, OpenCV AI Kit (OAK), and Other. Education includes Bachelor of Technology, Gati Shakti Vishwavidyalaya. AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and Classification.

Labeling Experience

Adaptive Traffic Control using Deep Reinforcement Learning (2025)

Don't discloseTextRLHF

In the adaptive traffic control project, the candidate designed a traffic signal optimization system using deep reinforcement learning. They trained/evaluated a reinforcement learning agent to reduce waiting time and improve throughput in a simulated environment. This constitutes AI training and evaluation for policy learning rather than data labeling. • Designed a SUMO-based traffic signal system • Implemented MAPPO for reinforcement learning control • Trained an agent to optimize traffic signal decisions • Reduced traffic waiting time and improved system throughput

2025 - 2025

Vessel Detection in SAR & EO Imagery (Sep 2025 – Nov 2025)

OpenCV AI Kit (OAK)Text

In the vessel detection project using SAR and EO imagery, the candidate built a maritime surveillance system with computer vision modeling. They evaluated model quality using mean average precision to assess detection performance. The work reflects supervised AI training and evaluation over labeled visual data or derived ground truth used for scoring. • Built a maritime surveillance system using Vision Transformers • Performed training/evaluation for vessel detection • Achieved mean average precision (MAP) of 0.88 • Applied the pipeline to SAR/EO imagery inputs

2025 - 2025

Intern – Central Railway Information Systems (CRIS), New Delhi (May 2025 – July 2025)

Don't discloseText

During the internship at Central Railway Information Systems, the candidate analyzed large-scale railway maintenance records using scalable data pipelines. They supported model development work related to predictive maintenance by deriving performance metrics from processed data. This activity aligns with AI training/evaluation on structured text-like tabular records rather than manual annotation. • Processed and analyzed 1 crore+ railway maintenance records • Used scalable data pipelines to prepare data for modeling • Developed RNN-based models (LSTM, GRU) and measured R² • Contributed to predictive maintenance strategies to reduce downtime

2025 - 2025

Track Defect Assessment – Metro Railway Kolkata (Mar 2024 – Apr 2024)

OtherTextClassification

For track defect assessment at Metro Railway Kolkata, the candidate applied probabilistic modeling and clustering techniques to support defect detection. The work indicates training and/or evaluation of AI models on defect-related data derived from observations. While explicit manual annotation is not described, the task is consistent with AI training over structured features or labels for defect classification. • Used Bayesian networks and clustering for defect detection • Supported identification of track defects for a metro rail system • Applied unsupervised/probabilistic techniques to improve detection signals • Built an AI-based assessment approach for maintenance use cases

2024 - 2024

Education

Work History