Shubhankar Tiwari | Software Engineer - Backend Systems & Applied AI

Active2026-Present

LLM Reliability Evaluation Platform

Production-grade system for measuring LLM behavioral consistency via Monte Carlo sampling. Benchmarked Qwen 2.5-7B across 240 inference calls — v2 grid sweep found temperature IS a lever, but top-p dominates 3.1×.

PythonFastAPINext.jsLLM+2

Completed2025

Indian Desi Multilingual LLM — Training Pipeline

End-to-end multilingual LLM training pipeline targeting Hindi/English code-switching. Dataset curation, LoRA fine-tuning, inference evaluation, and deployment packaging across 6 Kaggle notebooks.

PythonLLMNLPLoRA+2

Completed2021

Song Recommender System

ML-based workout song recommender using BPM and VADER sentiment analysis. Co-authored research with K-Means clustering on Billboard Top 100 to match songs to exercise intensity.

PythonMLNLPSentiment Analysis+2

Active2020-Present

Kaggle Portfolio

Notebooks Expert rank #2,441 / 59,663 — personal best #707. 34 notebooks, 11 datasets, 3 models, 1 competition entry. 10 bronze medals across ML, DL, NLP, Computer Vision, and regression.

Data ScienceMLDeep LearningNLP+1

Completed2025

Conservative Auto-Regeneration Policy for AI-Generated Financial Narratives

Production quality-gating system for AI-generated outputs in financial services — threshold optimisation, conservative AND escalation policy, priority-scored human review queue, and audit-grade provenance.

PythonNLPFinTechML+1