Yavar Khan
AI EngineerSoftware Engineer
ABOUT
Hi, I'm Yavar
During my 3 years at Accenture, I built a strong foundation in developing and maintaining reliable, production-grade systems at scale. I went on to earn an MS in Computer Science (AI/ML track) from SUNY Buffalo, focusing on the development and optimization of machine learning and deep learning models.
Since graduating, I have been expanding my skillset in Generative AI and putting it to work, building end-to-end AI applications using frameworks like LangChain, LangGraph, and PyTorch, from RAG pipelines to multi-agent architectures.
Currently, I am building WeatherWise.ai, a conversational multi-agent system designed to translate user queries into clear understanding backed by raw meteorological and atmospheric data and APIs. It delivers physics-grounded reasoning and actionable insights, helping users decode the complexity behind real-world weather phenomena.
Years · at Accenture
GPA · MS in CS
Projects · Shipped
What I Build
I bring the engineering expertise of deploying scalable, production-ready systems with the scientific depth required to build complex AI models.
GenAI Applications
Multi-agent architectures, RAG pipelines, and LLM-powered assistants with structured outputs, tool use, and production-grade observability.
AI/ML Systems
End-to-end machine learning pipelines from custom Transformers and CNNs to anomaly detection autoencoders and reinforcement learning agents.
Scalable Software
Production REST APIs, cloud-native deployments, and monitoring dashboards, built for high-concurrency traffic and operational reliability.
← Career Roadmap →
My Journey
From Engineering to Intelligence
5 milestones · 2016 → present
WeatherWise.ai
The GenAI Pivot
AI/ML Engineer Preparation
M.S. Computer Science (AI/ML Track)
GPA 3.87SUNY Buffalo
Accenture Soln Pvt Ltd
Pune, India
Bachelor of Technology, Information Technology
Amity University, India
WeatherWise.ai
The GenAI Pivot
AI/ML Engineer Preparation
M.S. Computer Science (AI/ML Track)
GPA 3.87SUNY Buffalo
Accenture Soln Pvt Ltd
Pune, India
Bachelor of Technology, Information Technology
Amity University, India
Featured Project
WeatherWise.ai
In collaboration with Ishita Srivastava
The purpose of computing is insight, not numbers.
— Richard Hamming
WeatherWise.ai is a conversational weather intelligence system. Ask anything from a simple forecast to a complex atmospheric science question, and get expert-level analysis grounded in real-time data and published research. No dashboards to navigate, no endpoints to query. Just a conversation that reasons like a meteorologist and cites like a researcher.
Ask Anything
Real questions. Real-time answers.
What’sthecurrentdewpointspreadinMiamiandshouldIexpectradiationfogbytomorrowmorning?
Live conditions, forecasts, and severe alerts via Vaisala Xweather MCP
Built with—LangGraph · Claude Sonnet 4 / 3.7 · MCP · LangSmith · Pydantic v2 · Python 3.11+ · GCP · Pinecone · uv
$engineering progress
WeatherWise.ai
Build Log.
Key milestones shipped so far and what's coming next.
12
Shipped
2
Building
3
Planned
Foundation
ShippedXweather MCP Integration
Real-time weather tools (forecasts, air quality, alerts, raster maps) connected via MCP with auth and tool filtering.
Sandboxed Python Calculator
AST-validated execution environment for math and statistical operations on weather data.
Projects
A selection of AI/ML and GenAI projects demonstrating end-to-end engineering
Generative AI
WeatherWise.ai
Agentic weather intelligence system with multi-agent architecture, corrective RAG with vision grounding, and MCP integration. Converts natural language into expert-level meteorological analysis grounded in NOAA, NASA, and AMS research.
Deep Research Pro
GitHubAI research co-pilot that orchestrates six specialized agents - planner, searcher, summarizer, follow-up evaluator, writer, and QA — across multi-wave research cycles with parallel execution and two-level caching (in-memory + SQLite). Generates 2K–5K word cited reports with inline references, exportable to Markdown, HTML, and PDF.
Context-Aware RAG Assistant
GitHubIntelligent RAG pipeline with FAISS vector indexing, runtime persona switching, and real-time Pushover alerts, deployed on Hugging Face.
Machine Learning & Deep Learning
Transformer Sentiment Analysis
GitHubCustom Transformer built from scratch in PyTorch with multi-head self-attention and Word2Vec embeddings, trained on 560K Yelp reviews.
VGG-16 vs ResNet-18
GitHubComprehensive comparison of VGG-16 and ResNet-18 on 30K images across 3 classes with hyperparameter tuning and data augmentation.
ThermoWatch-AE
GitHubConvolutional Autoencoder for unsupervised anomaly detection in machine temperature data using the Numenta Anomaly Benchmark.
EMNIST Character Recognition
GitHubCNN for 36-class handwritten character recognition (26 letters + 10 digits) trained on 100K EMNIST balanced samples.
GridWorld RL
GitHubSARSA vs n-Step Double Q-Learning comparison on custom GridWorld with Optuna-based hyperparameter optimization.
Speech Emotion Recognition
GitHub1D CNN for classifying 5 emotions from speech audio using MFCC features, data augmentation, and a Flask web interface.
Presentations
Paper presentation and code demo explaining how DeepMind's GraphCast, a Graph Neural Network, outperforms traditional numerical weather prediction models by predicting hundreds of global weather variables at 0.25° resolution for up to 10 days, surpassing ECMWF's HRES on 90.3% of verification targets.
Skills & Certifications
Programming & Backend
9 technologies
AI/ML & GenAI
16 technologies
Data & Retrieval
10 technologies
Cloud & Deployment
12 technologies
Tools & Platforms
10 technologies
Let's Build Something
Together
Whether it's an AI/ML role, a GenAI collaboration, or a conversation about intelligent systems — I'd love to hear from you.
yavarkhan1997@gmail.com
Email me directly
github.com/yavar29
Check my projects
in/yavar-khan29
Connect on LinkedIn
Download Resume
PDF • Updated 2026
Send a Message
I'll get back to you within 24 hours.