Available for full-time roles
saaz.dev@portfolio:~$ cat about.md

Building AI systems that ship.

I work at the intersection of machine learning, backend engineering, and product thinking. My recent work spans RAG systems, model governance tooling, fraud detection MLOps, and efficient computer vision pipelines.

35
public repositories

An active GitHub footprint across ML, backend, tooling, and experiments.

8.37
CGPA / 10

B.Tech in Computer Science with Data Science specialization at SRM IST.

5+
flagship builds

From governed AI systems to medical imaging and fraud pipelines.

$ summary

What I do best

  • AI product building: taking LLM or ML ideas from notebook stage to usable apps.
  • Retrieval and evaluation: building RAG flows with FAISS, sentence transformers, and measurable retrieval quality.
  • Reliable ML delivery: adding CI/CD, experiment tracking, testing, and monitoring around models.
  • Full-stack execution: shipping interfaces and APIs with FastAPI, Node.js, React, and Streamlit.
$ stack

Core stack

Python JavaScript PyTorch TensorFlow FastAPI React FAISS Docker GitHub Actions SQL / SQLite MongoDB AWS / GCP / Azure OpenAI
$ selected_highlights

Recent build highlights

  • Fraud Detection MLOps Pipeline: XGBoost pipeline on 590k IEEE-CIS transactions with ROC-AUC 0.93, CI, monitoring, and API serving.
  • Model Card Auditor: an AI governance tool focused on model-card quality, compliance checks, and developer-friendly audit workflows.
  • MemoryPal: a study assistant that converts notes into summaries, flashcards, and question answering with spaced repetition.
  • SA-EffNetB3: an attention-guided diabetic retinopathy model with 95.77% validation accuracy and edge-friendly deployment goals.
$ education

Academic track

SRM Institute of Science and Technology, Chennai

B.Tech in Computer Science with specialization in Data Science, 2022-2026.

  • Relevant areas: machine learning, deep learning, DSA, DBMS, DevOps.
  • Why it matters: I like pairing theory with production constraints instead of treating them as separate worlds.
  • Working style: fast iteration, measurable results, and clean handoff-ready implementation.
$ next

Where I can contribute immediately

Applied AI teams ML platform work Backend + API engineering Evaluation and observability Product-minded experimentation