My Melody from Sanrio
Grumpy Cat looking unimpressed
Cinnamoroll from Sanrio
The cat reading a newspaper meme
Chococat from Sanrio
The teary-eyed cat meme
Sania Gurung at an AI project exhibition
At an AI project exhibition.

Hi, I’m Sania.

Computer Science student.
AI & ML specialization.
Delhi-NCR, India.

I’m studying Computer Science & Engineering at Amity University, specializing in AI and machine learning. I graduate in 2027.

I build projects around AI agents, machine learning pipelines, and backend services. I’m interested in how the parts fit together: the model, the data, the API, and everything needed to make them work.

Most recently, I interned at Winniio / LifeAtlas, working on ML-based threat detection and dependency vulnerability scanning.

Selected work
  • A local LLM agent platform with tracing and automated evaluations.
  • A pipeline that trains a classifier on daily news headlines.
  • Computer vision and text classification, presented at AI Summit 2026.
  • A notification service with Kafka, Redis, and Spring Boot.

Explore the code on GitHub. Project details are below.

Projects

2024 — 2026

Selected projects in AI, machine learning, and backend engineering. Each includes source code and implementation notes.

A fully local LLM agent platform. It traces model calls, caches repeated requests, and runs automated quality and safety evaluations on each code push.

Ollama / FastAPI / Redis / Celery / Grafana

Implementation notes
  • Asynchronous task processing with Redis and Celery.
  • Grafana dashboards for latency, cache hit rate, and request history.
  • CI blocks deployment when safety-refusal evaluations fail.

An automated pipeline that collects BBC headlines, fine-tunes DistilBERT, and classifies articles as news, sport, or entertainment.

ScrapeWeak-labelFine-tuneServe

Airflow / DistilBERT / MLflow / DVC / FastAPI

Implementation notes
  • Dataset versioning with DVC and experiment tracking in MLflow.
  • Predictions served through a FastAPI endpoint.
  • The pipeline runs locally, without a cloud account.

An intelligence perception system combining object detection with a text-based threat classifier. It generates structured JSON reports for downstream systems.

Presented at AI Summit 2026, Bharat Mandapam, New Delhi.

TensorFlow / Computer Vision / Python

Implementation notes
  • Object detection for persons, vehicles, and aircraft.
  • A text classifier for threat-related inputs.
  • Structured output that can be consumed by other layers of the system.

A notification microservice that routes email, SMS, and push messages through Kafka, with deduplication, rate limiting, and retry handling.

Java / Spring Boot / Kafka / Redis / Supabase

Implementation notes
  • Redis-based deduplication and per-user sliding-window rate limiting.
  • Exponential-backoff retries and a dead-letter queue.
  • A REST API, Picocli CLI, Docker Compose setup, and Postman collection for API testing.

All repositories 

Experience

Full résumé 

Winniio / LifeAtlas

May — Jul 2026

AI Engineering Intern · Remote, Sweden

  • Designed and trained an ML model to classify security vulnerabilities by risk severity.
  • Built a dependency vulnerability scanner using CVE and OSV data, along with severity trend visualizations.

Stakrid

Aug — Dec 2025

AI Engineer Intern · Noida, India

  • Developed face detection for rider identity verification with real-time image processing.
  • Built a financial workflow module to flag transaction discrepancies using rules and anomaly detection. Also contributed to code reviews.

Background

I’m pursuing a B.Tech in Computer Science & Engineering at Amity University, with an AI & ML specialization, from 2023 to 2027.

My projects have taken me from training models to building the services around them. I like understanding both sides, and seeing a system through from input to output.

This page brings together that work so far.

Technical toolkit

AI & machine learning
Python, TensorFlow, Scikit-learn, Pandas, NumPy, Ollama
Backend & data
Java, FastAPI, Spring Boot, SQL, Redis, Kafka, Celery, PostgreSQL
Pipelines & infrastructure
Airflow, MLflow, DVC, Docker, GitHub Actions, Grafana

Additional coursework

  • Machine Learning SpecializationStanford Online · 2026
  • Data Structures & AlgorithmsCoding Ninjas · 2024–2025
  • Java Spring Framework 6 with Spring Boot 3Udemy · 2025

Get in touch

LinkedIn 

For engineering opportunities, project discussions, or questions about my work, please get in touch:

Opens your email app with your message. Nothing is posted on this site.

Header image credits

My Melody, Cinnamoroll, and Chococat artwork: Sanrio. Cat meme images sourced via Imgflip. All images belong to their respective owners.