Chaman Vashishth

AI/ML Engineer · ML Systems · AI Research

Hathras, Uttar Pradesh, India

@chaman_vashishth

GitHub LinkedIn

0 followers

🎓 Bachelor of Vocation in Internet of Things at Dayalbagh Educational Institute · Graduating 2028

About

AI/ML engineer building machine learning systems from first principles, local-first AI software, and applied ML applications. Has contributed to open-source projects and built projects spanning neural networks, recommendation systems, and local-first AI/quantum tools.

Experience

AI & Machine Learning Intern

InAmigos Foundation

Aug 2026 – Sep 2026

Built an AI Resume Optimizer as part of an applied ML internship, working across practical ML development and experimentation. Worked with iterative model experimentation and evaluation for an applied AI/ML workflow.
AI/ML Engineer Intern

Akiyam Solutions Pvt. Ltd.

Jul 2026 – Aug 2026

Worked across data preprocessing, model experimentation, model evaluation, and iterative ML development. Used Git-based collaboration and code-review workflows.
Open Source Contributor

Freelance

2026 – 2026

Contributed to GSoC Org Finder, ML-CaPsule, and Student Notes App across responsive UI, offline functionality, analytics, and productivity features.

Skills

<3 months

C SQL Shap NumPy React Scipy Python Pandas Streamlit JavaScript TypeScript scikit-learn Neural Networks Transformer architectures API Integration Understanding

Tools / apps / platforms

<3 months

Git Vite Linux Docker GitHub Pytest

Projects

Movie Recommendation Engine ↗

Python · NumPy · SciPy · Pandas · Streamlit · Docker

Built collaborative filtering with sparse matrices, SVD initialization, user/item bias, and analytically solved ALS on MovieLens 1M. Achieved 0.865 RMSE, 0.682 MAE, 0.420 NDCG@10, and 0.680 Hit Rate@10 with cold-start fallback and explainable Top-10 recommendations.
Neural Network from Scratch ↗

Python · NumPy · Matplotlib · Requests

Implemented a 235,146-parameter network from first principles with manual forward/backprop, BatchNorm, Dropout, stable Softmax/Cross-Entropy, and Adam. Trained on 60,000 MNIST images; achieved 97.4% accuracy and 97.0% macro F1.
Quantum Laboratory (AIQCP) ↗

React 19 · TypeScript 5.9 · Vite 7 · Node.js 22.12

Built a local-first quantum-computing laboratory connecting curriculum, circuit simulation, adaptive practice, grounded tutoring, experimentation, and research workflows. Implemented statevector simulation, measurement/sampling, gates, visualizations, quantum DSL, noise, density-matrix channels, partial trace, purity/entropy, adaptive learning, and English/Hindi localization.
ARIA ↗

Python 3.11+ · custom Tensor/autodiff · neural layers · byte tokenizer · Transformer SLM · local training · pytest

Building a software-only local-first AI stack around an ARIA-owned trainable neural brain rather than a hosted model API. Implemented tensor ops/reverse-mode autodiff, trainable neural layers, deterministic byte tokenizer, trainable LM core, experimental decoder Transformer SLM, and local training foundation.

Courses & certifications

Introduction to Generative AI · Google Cloud · 2026
Python for Data Science · IBM · 2026
Machine Learning Specialization · DeepLearning.AI · 2026

🏆 Achievements & awards

OpenAI × Outskill Hackathon 2026 — Top 10% · 2026

Placed in the top 1,000 among 10,000+ applicants.
Amazon ML Summer School 2026 — Qualified · 2026

Qualified for the Selection Test after resume + SOP screening.

Explore the Broxer community

💬 Q&A Ask about roles, salaries and interviews — and help others. 👥 Circles Join discussions on skills, companies and industries. 📰 Feed Follow updates and insights from the community.