Moiz's Photo
👋 Hello, I'm

Syed Moiz Ali

ML Engineer & Infrastructure Architect

Principal AI/ML Engineer and Systems Architect with nearly a decade of experience designing, scaling, and deploying production-grade ML ecosystems and cloud infrastructures. Expert in Generative AI, LLMOps, RAG Architectures, and distributed systems. IIT Kanpur alumni.

Download Resume

About

Principal AI/ML Engineer and Systems Architect with nearly a decade of experience designing, scaling, and deploying production-grade ML ecosystems and cloud infrastructures. Merges a rigorous academic foundation in Operations Research (IIT Kanpur) with deep expertise in Generative AI, LLMOps, and distributed systems. Proven track record in Agentic Engineering, RAG Architectures, and high-availability enterprise SaaS platforms.

Core Expertise

Cloud & Infrastructure
AWS Terraform Kubernetes Docker
Backend & Data
Python Node.js TypeScript PostgreSQL Redis
ML & AI
PyTorch TensorFlow MLflow Spark

Experience

  • Snorkel AI: Engineered high-fidelity programmatic labelling datasets advancing foundational model safety and evaluation consistency across AI benchmarks.
  • PaperBench: Integrated robust test suites and validation modules into the ML framework, improving validation metric accuracy for automated code generation environments.
  • Dreamflow Buildathon 2025: Architected AI agents and integrated emergent Generative AI frameworks, accelerating concept-to-deployment execution for autonomous workflows.

  • OutsourceGuard SaaS Platform: Orchestrated end-to-end delivery using Agentic Engineering to rapidly scaffold application layers around core AI engines, accelerating time-to-market.
  • Vendor Risk Modeling: Reduced manual document review time by 40% by architecting automated risk scoring systems using fine-tuned Hugging Face Transformers.
  • Infrastructure: Ensured 99.9% uptime (CloudWatch validated) by provisioning AWS environments with custom domain routing and automated GitHub Actions CI/CD pipelines.

  • Time-Series Forecasting: Improved industrial time-series forecast reliability with double-digit gains by designing ensemble architectures integrating gradient boosting and LSTM networks.
  • RAG & NLP: Enhanced domain-specific conversational AI performance by 18% via custom RAG architectures and advanced prompt engineering techniques.
  • MLOps & Cloud: Reduced deployment time by 60% and pipeline failures by 85% by migrating to Terraform and containerised services on AWS ECS and SageMaker.
  • Document Classification: Achieved 94% precision by engineering scalable CNNs with Optuna hyperparameter tuning and SHAP explainability frameworks.
  • API Reliability: Maintained 99.2% API availability during 300% traffic spikes by orchestrating FastAPI microservices across Docker and Kubernetes clusters.

  • Database Optimisation: Reduced average query execution times by 32% through comprehensive schema refactoring and advanced B-tree indexing in PostgreSQL.
  • Data Pipelines: Eliminated critical reporting discrepancies and pipeline outages (zero downstream alert triggers over consecutive quarters) by leveraging Apache Spark for automated, high-volume data processing workflows.

Project Details

Orchestrated end-to-end delivery of the OutsourceGuard enterprise SaaS platform using Agentic Engineering to rapidly scaffold application layers around core AI engines. Architected Vendor Risk Modeling systems using fine-tuned Hugging Face Transformers for automated risk scoring, reducing manual document review time by 40%. Provisioned AWS infrastructure with GitHub Actions CI/CD, achieving 99.9% uptime validated via CloudWatch.

Deployed custom Retrieval-Augmented Generation (RAG) architectures with advanced prompt engineering techniques, achieving an 18% increase in internal NLP accuracy benchmarks for domain-specific conversational AI systems.

Designed and deployed complex ensemble architectures integrating gradient boosting (XGBoost, LightGBM) and LSTM networks for industrial time-series forecasting, achieving double-digit performance gains over baseline models.

Engineered scalable CNN pipelines integrated with Optuna hyperparameter tuning and SHAP explainability frameworks, elevating document classification accuracy to a 94% precision rate.

Orchestrated FastAPI microservices across Docker and Kubernetes clusters, scaling backend model inference to maintain 99.2% API availability during 300% traffic spikes.

Education

  • Institute: Indian Institute of Technology, Kanpur
  • Department: Department of Management Sciences
  • Major: Production & Operations Management
  • CGPA: 8.0
  • Year: 2009-2011
  • Location: Kanpur, Uttar Pradesh, India

  • Institute: RITEE, CSVTU
  • Department: Electronics & Telecommunication Engineering
  • Major Project: Implementation of Image Processing and Image Enhancement using MATLAB.
  • CGPA: 8.5
  • Year: 2005-2009
  • Location: Raipur, Chhattisgarh, India

Certifications

  • Date of Completion: June 2021
  • Certification Provider: Stanford Online
  • View Certification

  • Date of Completion: April 2021
  • Certification Provider: DeepLearning.ai
  • View Certification

  • Date of Completion: November 2022
  • Certification Provider: DeepLearning.ai
  • View Certification

Research Experience & Publications

  • Institution: Sultan Qaboos University
  • Year: 2014-2015
  • Undertook a project titled: "Mediator-based order acceptance decision system under the make-to-order company."
  • Worked under the guidance of Dr. Sujan Piya, focusing on improving order acceptance mechanisms.

  • Company: Central UP Gas Limited
  • Year: 2010
  • Market Analysis: Led an initiative to examine the market potential for natural gas in the Rania & Jainpur Industrial Areas.
  • Formulated strategies to establish Piped Natural Gas service stations, enhancing the distribution network.

  • Sharma, R. R. K., & Ali, S. M. (2017). Reducing a Lot Sizing Problem with Set up, Production, Shortage and Inventory Costs to Lot Sizing Problem with Set up, Production and Inventory Costs. American Journal of Operations Research, 7, 282-284. Link
  • Ali, S. M., Sharma, R.R.K., & Gupta, O.K. (2015). Lagrangian Relaxation Procedure for the Capacitated Dynamic Lot Sizing Problem. AIMS International Conference on Management. Link
  • Syed, M. A., & Sharif. (2012). Aggregate Planning for Semi-finished goods under make-to-stock environment. International Journal of Advances in Management, Technology & Engineering Sciences, 1(8(I)), 104-107.
  • Sharif, & Syed, M.A. (2012). Procurement Policies & Inventory Management System in Manufacturing and Service Settings: An Optimization Framework. International Journal of Business, Management & Social Sciences, 1(9), 27-32.
Send an email to communicate.

Get in Touch

Syed Moiz Ali

9984673534

moizeali@gmail.com

Raipur, India

Copyright © 2026 Moiz Ali