Bikash Chandra Sahoo

Bikash Chandra Sahoo

About

Detail

Software Developer. Python. Django | Research. ML | Big Data Analytics, M.Tech
Hyderabad, Telangana, India
Contact Bikash regarding: 
work
Full-time jobs
Starting at USD25/hour
Flexible work
Starting at USD25/hour
id_card
Internships
Starting at USD1.2k/month
groups
Networking

Timeline

work
Job
school
Education
folder
Project
auto_stories
Publication

Résumé

Jobs verified_user 0% verified
  • Outlier
    Handshake AI
    Freelance
    Outlier, Handshake AI
    Jul 2024 - Current (2 years 5 months)
    Review AI generated code and content by grading it based on different dimensions.
  • S
    Software Developer
    Sumprio Technologies Pvt Ltd
    Dec 2025 - Mar 2026 (4 months)
    • Designed and maintained scalable APIs for a Django-based CRM application using Python.
    • Optimized API queries and implemented structured unit tests for the entire application, ensuring code reliability, improved test coverage , and robust error handling.
    • Maintained detailed feature design and technical documentation using Confluence, ensuring clear communication of requirements and system workflows.
  • S
    Software Developer Trainee
    Sumprio Technologies Pvt Ltd
    May 2025 - Nov 2025 (7 months)
    • Understood application design and workflow.
    • Conducted performance and stress testing using Apache JMeter, identifying bottlenecks and optimizing server response time.
  • Z
    Data Science and ML Intern
    Zaalima Development
    Apr 2025 - Jul 2025 (4 months)
    • Developed a Fraud Detection System leveraging machine learning models (Random Forest, XGBoost) to achieve 95% accuracy in identifying fraudulent transactions.
    • Built an AI-powered Resume Screening System with spaCy, TF-IDF, and BERT integration, improving
    recruitment efficiency by automating resume-job description matching with a 92% accuracy rate.
  • C
    Data Science Research Intern
    California Institute of Genetics
    Nov 2024 - Jan 2025 (3 months)
    • Contributed research to develop an AI-driven approach for early Parkinson's disease diagnosis.
    • Collaborate with a multidisciplinary team to design and implement machine learning models aimed at identifying biomarkers of Parkinson's disease.
    • Analyze complex genetic and clinical datasets to uncover patterns and insights that inform model development.
  • CSIRNational Geophysical Research Institute
    Summer Trainee
    CSIRNational Geophysical Research Institute
    Jun 2018 - Aug 2018 (3 months)
    • Analyzed and image process seismic data using balanced contrast enhancement technique (BCET) in MATLAB.
    • Achieved a low MSE of 0.0014 and a CNR of 2.1, enhancing image detail preservation and contrast.
Education verified_user 0% verified
  • Vellore Institute of Technology
    Master of Technology - MTech, Computer Science with specialization in Big Data Analytics
    Vellore Institute of Technology
    Jan 2022 - Dec 2024 (3 years)
    Post graduation degree in Computer Science. The course consist of research thesis and publishing the same in peer reviewed journals. Gained experience in DevOps by utilizing Docker and Kubernetes in personal and academic projects.
  • A
    Bachelor of Engineering, Electrical, Electronics and Communications Engineering
    Army Institute of Technology College of EngineeringPune
    Jan 2015 - Dec 2019 (5 years)
  • Kendriya Vidyalaya
    Intermediate, Mathematics and Computer Science
    Kendriya Vidyalaya
    Jan 2014 - Dec 2015 (2 years)
Projects (professional or personal) verified_user 0% verified
  • T
    TinyShift
    Aug 2026 - Current (4 months)
    An AI based habit tracker that tracks habit and suggest to cope up if any habits are failed to commit on particular day by doing its smaller version. Developed using React.js.
  • SpotterLabs
    RouteLog(Professional)
    SpotterLabs
    Sep 2026
    An automated trip planner and ELD log generator for property-carrying drivers. The HOS rules engine behind it enforces every constraint a driver operates under — driving limits, shift windows, mandatory breaks, fuel intervals, and the 70-hour weekly cycle. Every plan comes with 24-hour log sheets in FMCSA § 395.8 format, ready for the road.
  • M
    Memoraq AI - Memo generator for decision making
    Jun 2026
    Helps in taking decision in a curated manner by analyzing the confusion the user have.
  • D
    Designing and curating interior designs
    May 2026 - Jul 2026 (3 months)
    Created interior design ideas using Nano Banana and replicating the same in home interior.
  • D
    Design a website for a boutique which has growing marketing.
    May 2026 - Jun 2026 (2 months)
    Design the sample frontend using html, css and javascript for a newly growing botique which wanted to expand the reach via personal website.
  • S
    Smart Outreach Tool: An AI-Powered Cold Email Generator
    Dec 2024
    Built a web-based tool to automate and personalize cold email generation using LLMs, with A/B testing and job-based customization to boost outreach effectiveness.
    • Extracts job details from career page URLs using Llama and formats data into structured JSON for targeted messaging.
    • Matches user portfolio links via ChromaDB vector store and integrates them into AI-generated emails.
    • Offers A/B testing with two tone styles—formal and conversational—to optimize engagement rates.
    • Leveraged full-stack development expertise to build end-to-end project architecture.
  • F
    Food delivery app dashboard
    Aug 2024
    Created a Power BI dashboard to track sales and derive insights from a food delivery app data.
  • S
    Search Engine Performance Optimization Using Enhanced Page Ranking Algorithm
    May 2024 - Jul 2024 (3 months)
    Built a smarter version of the PageRank algorithm to cut down computation time and improve ranking accuracy.
    • Achieved 45% faster computation using Stother-William’s optimization on a 260k+ node web graph
    • Improved convergence (only ~25 iterations) with better alignment to node centrality
    • Tested on real-world SNAP dataset, showing strong performance for scalable search and network analysis
  • E
    Epileptic Seizure Prediction Using Deep Learning and Continual Learning
    Jan 2024 - May 2024 (5 months)
    Built a deep learning framework to predict seizures in patients with temporal lobe epilepsy using EEG signals—aiming to improve early intervention and real-time alerts.
    • Preprocessed time-series EEG data using Butterworth-Wavelet filters (24.11 dB SNR) to reduce noise and enhance signal clarity.
    • Achieved 97.5% accuracy with GoogleNet through transfer learning; improved to 98.7% with Elastic Weight Consolidation (EWC) for continual learning.
    • Designed the model to retain past knowledge while learning new patterns—solving the problem of catastrophic forgetting.
  • H
    Harnessing Quantum Computing for Stock Market Prediction.
    Jul 2023 - Dec 2023 (6 months)
    Explored the use of Quantum Machine Learning (QML) to improve stock market prediction accuracy by simulating quantum linear regression models using Qiskit.
    • Preprocessed and encoded historical stock data as quantum states using Hadamard and Controlled-Z gates for superposition and entanglement.
    • Designed a quantum circuit with parameterized gates and trained it via quantum linear regression, achieving an MSE of just 1.91%.
    • Evaluated results using metrics like RMSE, MAE, and R² (0.68), demonstrating the potential of quantum computing in complex financial forecasting.
  • A
    A deep Learning Framework for Prediction of Cardiopulmonary Arrest.
    Sep 2022 - Jun 2023 (10 months)
    Built and evaluated multiple machine learning models to forecast cardiopulmonary arrest, with ANN achieving the highest accuracy of 94% on the Cleveland dataset.
    • Used two public health datasets (Cleveland & Framingham) with both numerical and categorical risk factors like cholesterol, BMI, and glucose.
    • Applied advanced techniques like Transfer Learning (ResNet18), MAML, and Layer-wise Relevance Propagation for model tuning and explainability.
    • Converted tabular data to heatmap images for CNN-based models, and integrated gamma prime fibrinogen as a novel predictive attribute
Publications verified_user 0% verified
  • IEEE
    Robust Seizure Prediction Model Using EEG Signal for Temporal Lobe Epilepsy Leveraging Deep Learning and Continual Lea
    IEEE
    Aug 2025
    Developed a deep learning–based seizure prediction model for temporal lobe epilepsy using EEG signals. Leveraging GoogLeNet with continual learning via Elastic Weight Consolidation (EWC) and a custom learning rate scheduler, the model achieved 98.7% accuracy by improving 1.2% over the baseline and hence enhancing adaptability, preventing catastrophic forgetting, and optimizing temporal pattern recognition in noisy time-domain data.
  • Springer Nature
    Decoding Market Dynamics: Variational Quantum Circuit in Stock Prediction
    Springer Nature
    Jun 2025
    Explored the powerful potential of Quantum Machine Learning (QML) to predict trends in the ever-evolving financial markets. Using a quantum variational circuit classifier with optimizers like Real Amplitude and EfficientSU2, the model was trained on a cryptocurrency-based stock dataset.
  • E
    A Deep Learning Framework for the Prediction of Cardiopulmonary Arrest
    European Alliance Innovation EAI on Pervasive Health and Technology
    Mar 2024
    Conducted a comparative study on predictive models for cardiopulmonary arrest and stroke, analyzing key health factors to improve early detection. Evaluated ANN, Transfer Learning, MAML, and LRP, with ANN achieving the highest accuracy of 94%, highlighting its potential for proactive diagnosis and timely medical intervention.
  • I
    American Sign Language Translator and Calling Device.
    IPASJ International Journal of Electronics and Communication
    May 2019
    Used raspberry pie, flex sensor as the major component of the module. Tabulated different resistance values based on gesture obtained from flex sensor. Using python, annotated the gesture and converted it into text and voice.