Open to work

Sooraj Kakkanatt Manoj

I build ML pipelines, RAG systems, and data infrastructure — MS in Data Science from RIT, now looking for full-time roles in data science and AI/ML engineering.

ContactProjects
Sooraj Kakkanatt Manoj

Skills

Programming & Databases

PythonSQLRJavaJavaScriptPostgreSQLMySQLSnowflakeMongoDB

Machine Learning & Deep Learning

PyTorchTensorFlowKerasLangChainCUDANLPHuggingFace

MLOps, Cloud & Deployment

AWSAzure MLGitDockerKubernetesFastAPIFlask

Data Engineering & Visualization

Apache SparkApache AirflowHadoopTableauPower BIJupyterQGISSPSS

Projects

Research

Bias in AI-Generated Travel Narratives: Evaluating LLMs in Countering Misconceptions

Anshuman Mohapatra, Ashique Khudabukhsh, Bhaskar Sreenivas Pavan Akkena, Sooraj Kakkanatt Manoj, Travis Desell

Rochester Institute of Technology, Department of Software Engineering · 2026

Investigates how bias manifests in AI-generated travel narratives, treating LLMs not as objects of evaluation but as instruments for revealing representational, cultural, and evaluative bias. Compares counterspeech generated by LLaMA 3.2, Gemma 2, GPT-2, and Sarvam across human and automated toxicity assessments, showing that automated scoring consistently underestimates implicit harm and produces unstable model rankings — underscoring the need for context-aware, human-grounded evaluation of generative AI systems.

Read paper →

Analyzing Social Media Responses Across Various Languages Based On RG Kar Incident

Sooraj Kakkanatt Manoj, Yogesh Kuchimanchi

Rochester Institute of Technology, Department of Data Science · 2026

Explores the multilingual dynamics of public discourse surrounding the RG Kar Hospital incident using YouTube comments in five languages. Applies VADER sentiment analysis and LDA topic modeling to uncover recurring themes, revealing an unexpected dominance of political discourse and administrative-failure criticism over the initially expected focus on misogyny — highlighting the power of public sentiment analysis in surfacing societal concerns.

Read paper →

Self Projects

A filterable gallery across AI/ML, data, and infrastructure work.

AI-Powered YouTube Q&A Assistant

Fully local RAG pipeline for YouTube Q&A using FastAPI, LangChain, and Ollama (llama3 + nomic-embed-text) — indexing transcripts in under 2s and answering in 8–12s. Session-scoped semantic search with FAISS retrieval and per-video chat memory, plus a pytest suite (13 tests, 71% coverage) with mocked LLM calls for CI reliability.

FastAPILangChainOllamaFAISSRAG
Code →

Dockerized MLOps API with FastAPI

Containerized a TF-IDF + Logistic Regression sentiment classifier behind a FastAPI service with Pydantic-validated schemas and auto-generated OpenAPI docs. CI via GitHub Actions runs the pytest suite and validates the Docker build on every push/PR.

FastAPIDockerscikit-learnGitHub Actions
Code →

Kafka-Based Credit Card Transaction Processing System

Real-time + batch transaction pipeline on a Kafka-based Lambda Architecture, streaming 400+ transactions through fraud checks (geolocation mismatch, credit-utilization limits) before batch-layer reconciliation. Stream-layer validation flags 2.5% of transactions for anomalies, then pushes finalized balances/scores back to MySQL.

PythonKafkaMySQLLambda Architecture
Code →

Healthcare Data Normalization with PySpark

Spark ETL pipeline normalizing flat legacy healthcare visit data into a 10-dimension + 1-fact snowflake schema, with automated referential integrity checks across all foreign keys and Tableau dashboards for insight generation.

PySparkMongoDBTableau
Code →

YouTube Content Creator Recommendation System

A recommendation system for YouTube content creators.

PythonRecommendation Systems
Code →

Education

  • Rochester Institute of Technology

    MS in Data Science

    May 2026

  • PSG College of Arts and Science

    BSc in Computer Science with Data Analytics

    May 2023

Certifications

  • Six Sigma White Belt

    Council for Six Sigma Certification

    Jul 2026

  • Microsoft Azure AI Fundamentals (AI-900)

    Coursera / Microsoft

    May 2025

  • Diploma in Programming and Data Science

    IIT Madras

    Aug 2022

Experience

  1. TEACHING & RESEARCH · Aug 2025 – Apr 2026

    Graduate Teaching Assistant — Department of Computer Science

    Rochester Institute of Technology · Rochester, NY

    Mentored 25 students per semester in advanced server programming, backend system logic, and automation concepts, grading 750+ weekly script and data-pipeline assignments across 2 semesters with code reviews. Held 120+ hours of office hours advising 60–90 students on debugging and production-ready code.

  2. INTERNSHIP · Jan 2022 – Jul 2024

    Intern: Analysis & Design of Commissions for Gig Workers

    One Integral Technologies Pvt. Ltd. · Chennai, India

    Designed and optimized performance-based pay models using Digital Twin simulations across 50+ data points, improving worker engagement by 15%. Built a React.js analytics dashboard turning data from 10+ sources into visualizations covering 500+ gigs, and integrated it with the backend Digital Twin system via MongoDB for predictive incentive analysis.

Get in touch

Actively looking for full-time roles in data science and AI/ML engineering — also open to research collaborations and freelance work.

Currently on F-1 OPT, eligible for the STEM extension — authorized to work in the U.S. with no sponsorship required for up to 3 years.