Headline

AI/ML Engineer | MLE Consultant | AI Annotator | RLHF Expert | Workflow Automation | AI agents

About

Hello, I'm Aryanil Dey. Thank you for visiting my profile, if you are in a hurry, you are in the right section. I’ve worked across healthcare and SaaS, taking models from messy raw data to monitored production APIs: improving clinical ROC‑AUC from 0.76→0.88, reducing false negatives, cutting p95 latency from 450 ms→180 ms, and stabilizing ETL pipelines to 99% success so downstream analytics and products always have trustworthy signals. I also design AI infrastructure around MLflow/DVC, experiment tracking, drift monitoring, and rollback playbooks so models stay robust after deployment. On the GenAI side, I work on LLM powered tools, RLHF style evaluation, and AI annotation probing model behavior, designing rubrics, and turning those insights into better prompts, datasets, and guardrails. I co‑authored “Lung cancer type prediction using LightGBM machine learning model” and write about topics like the evolution of language representation (from BoW to GPT) and search‑engine system design. Toolbox: Languages: Python, R, SQL, SAS, HTML, Git Machine & Deep Learning: supervised learning, NLP, time series, CNN/RNN/LSTM, autoencoders, GANs, model calibration, SHAP, imbalanced learning AI Infrastructure & MLOps: Pandas, NumPy, scikit‑learn, XGBoost, CatBoost, MLflow, DVC, REST APIs, microservices, Docker, testing/monitoring pipelines Cloud & Frameworks: Flask, Django, AWS EC2, AWS S3, Amazon SageMaker, Amazon Bedrock Analytics & Visualization: Power BI, Excel, Plotly, matplotlib, Seaborn; experiment design, A/B testing, forecasting Beyond Data: I love exploring new places, music, cooking, and cricket. A cricket fan through and through, I’ve also developed a knack for sports analysis. I’m equally passionate about pop culture from movies to comics and always find ways to blend creativity into my work. 🤝 Let’s Collaborate: If you share a passion for data or want to collaborate on data-driven projects, feel free to reach out!

Location

Bengaluru

AI stack

claude code

hugging face

kiro

langchain

perplexity

Amazon Q Developer

Work experience

AI/ML Engineer at Biodesign Innovation Labs, Nov 2024 - Present · 1 yr 10 mos, Bengaluru, Karnataka, India

RLHF Expert at Invisible Technologies, Apr 2026 - Jun 2026 · 3 mos, Training foundational model on data-driven insights and evaluating the models response on human reasoning and business operations and guardrails.

Machine Learning Specialist at Invisible Technologies, Dec 2025 - Mar 2026 · 4 mos, Solving applied ML tasks: iterating on problems that test and improve model reasoning, from baseline solutions to optimized workflows. Evaluate scientific writing: compare research abstracts and introductions using review rubrics (e.g., NeurIPS guidelines) for clarity, correctness, and impact. Debug advanced model behavior : probe LLMs on optimization, fairness, regularization, transformers, and applied NLP documenting failure modes.

Data Science With GEN AI at Innomatics Research Labs, Sep 2024 - Nov 2024 · 3 mos, Hyderabad, Telangana, India

Data Management Consultant at Sal, Aug 2024 - Sep 2024 · 2 mos, On-site

Data Analyst at Sal, Apr 2024 - Aug 2024 · 5 mos, Hybrid

Machine Learning Engineer at Social+, Dec 2023 - Mar 2024 · 4 mos, Greater Kolkata Area, Shipped NLP tone/sentiment classifier on 50k labeled tweets using scikit‐learn/XGBoost, achieving 92.4% accuracy and macro‐F1 0.90; reduced manual moderation workload ~40%. Implemented Selenium + API ingestion; increased data throughput 3.2× and reduced labeling cycle time 25% via retry/backoff and schema checks. Packaged model as a REST service with telemetry and A/B hooks, accelerating feedback loops for iteration.

Education

Master's degree, Data Science and Analytics, Jain (Deemed-to-be University)

Biotechnology, Bioinformatics, Garden City University

Public links

github: https://github.com/AryanilAD

medium: https://medium.com/@aryanilofficial26

youtube.com: https://youtube.com/@aryanildey