Production AI systems

Nikhil Shingadiya

I build

Machine Learning Engineer focused on reliable LLM applications, multi-agent orchestration, retrieval systems, and scalable AI infrastructure.
5+ yearsAI/ML experience
GenAILLMs, RAG & agents
End-to-endmodel-to-cloud delivery
Professional profile

About me

Machine Learning Engineer

Building production-grade AI systems from model to infrastructure.

I specialize in Generative AI, LLMs, RAG, and multi-agent architectures, with experience designing autonomous workflows, resilient model routing, secure execution services, and scalable AI backends. My work spans Python and FastAPI services, distributed workers, vector databases, model serving, cloud infrastructure, and reliable end-to-end machine-learning delivery.

Current companyRishabh Software
SpecializationGenAI, LLMs, RAG & agents
EngineeringPython, FastAPI & distributed systems
LocationAhmedabad, Gujarat, India
EducationB.E. Computer Engineering · 8.62/10
LanguagesEnglish, Hindi & Gujarati
Career

Professional experience

Selected work delivering production AI platforms, data products, and machine-learning systems.

Rishabh Software

Machine Learning Engineer
July 2025 – Present

Ahmedabad, Gujarat, India

GO AI
  • Architected a production multi-agent AI system with 5+ specialized agents for autonomous planning, execution, validation, and output generation, reducing task completion time by 45%.
  • Engineered intelligent LLM routing and failover across multiple models, achieving 98% service availability.
  • Built Data George, an orchestration layer that discovers, validates, and normalizes data from 20+ sources, including FRED, World Bank, and arXiv.
  • Developed a secure FastAPI, Redis, ARQ, and Docker code-execution service processing 10,000+ isolated Python jobs per day with automated error detection and self-repair.
  • Built a multi-surface engine for interactive HTML, visualizations, diagrams, and documentation, reducing manual content-generation effort by 70%.
BI Agent
  • Developed a multi-user business-intelligence platform that converts natural-language questions into SQL, visualizations, and analytical insights with role-based access control.
  • Achieved 85% NL-to-SQL accuracy and reduced token costs by 30% using retrieval-score optimization.
  • Integrated Intel OPEA for schema extraction and built a LangGraph workflow for contextual FAQ generation.
  • Engineered asynchronous processing with FastAPI, Celery, and Redis; used Milvus and MinIO for private, self-hosted enterprise deployments.
  • Built an LLM-powered ECharts engine that converts SQL results into interactive business visualizations.

Unlimited WP

AI/ML Lead Developer
Dec 2023 – July 2025

Ahmedabad, Gujarat, India

WEAM AI
  • Engineered an AI ecosystem for enterprise onboarding and collaborative document intelligence.
  • Designed a modular layered architecture with FastAPI, Celery, and Redis to improve maintainability and scale.
  • Built a specialized RAG system and LangChain agents for grounded answers and multi-step document guidance.
  • Integrated Pinecone and Qdrant for vector retrieval and used Ray to maximize GPU utilization.
  • Integrated OpenAI, Hugging Face, Gemini, and Anthropic models across the platform.
  • Deployed the backend on AWS EC2 with Docker, CI/CD, proactive monitoring, and CloudWatch-based worker scaling.

GARUDA AI

Unlimited WP — AI-Driven Blog Writing Engine

  • Developed an AI-driven blog engine that creates SEO-optimized content to improve website rankings.
  • Used Google Keyword Planner and Ahrefs data to identify high-impact keywords and relevant source material.
  • Automated optimized title, keyword-tag, and long-form content generation using Generative AI and LLMs.

F(x) Data Labs PVT LTD

Machine Learning Engineer
Jan 2022 – Dec 2023

Ahmedabad, Gujarat, India

GMR Group — MCP Prediction
  • Built a Random Forest model to forecast Day-Ahead Market Clearing Price with 70% accuracy (±6%).
  • Combined IEX Market, weather, and Google Trends data and designed confidence metrics for each time block.
  • Used Monte Carlo simulation to demonstrate an 11% profitability improvement over random trading strategies.
Wowsly Video Rendering
  • Developed a multi-user FastAPI and OpenCV service for CSV-driven video editing, S3 storage, and progress tracking.
  • Designed an asynchronous AWS architecture with Lambda, EC2, S3, Celery, Redis, Docker Compose, and Flower.

Education & Certifications

Bachelor of Engineering in Computer Engineering

2018 – 2022
CGPA: 8.62/10

L.D. College of Engineering (GTU), Ahmedabad, Gujarat

Training: Statistics with Python — University of Michigan (Coursera); Mathematics for Machine Learning — Imperial College London; Computational Thinking and Data Science — MIT.

Selected work

Personal projects

Applied machine-learning and data products built to explore real user problems, deployment patterns, and analytical workflows.

Sumquiry AI PDF assistant Generative AI

Sumquiry PDF Assistant

Interactive PDF summarization and question answering with semantic retrieval and grounded responses.

RAGLangChainStreamlit
View case study
Customer churn prediction interface Machine Learning

Customer Churn Prediction

End-to-end classification workflow for predicting repeat food orders from customer behavior and feedback.

Random ForestStatisticsFlask
View case study
Movie recommendation system Recommendation

Movie Recommendation System

Content-aware recommendation experience that helps users discover relevant titles from preference signals.

PythonSimilarityData Pipelines
View case study
Electronics sales analysis dashboard Analytics

Electronics Sales Analysis

Exploratory analysis and visual storytelling focused on sales patterns, outliers, and commercial trends.

PandasEDAVisualization
View case study
YouTube analytics dashboard Data Product

YouTube Analytics Dashboard

Interactive dashboard for exploring content performance through responsive Plotly visualizations.

DashPlotlyPython
View case study
Robot simulation visualization Simulation

Robot Movement Simulation

Computational simulation exploring robot movement, room cleaning strategies, and performance over time.

PythonAlgorithmsSimulation
View case study
Continuous learning

Credentials

Formal training that strengthened my statistical reasoning and mathematical foundation for machine learning.

Statistics with Python certificate
University of Michigan · Coursera

Statistics with Python

Statistical inference, confidence intervals, hypothesis testing, regression, and practical visualization.

View credential
Mathematics for Machine Learning certificate
Imperial College London · Coursera

Mathematics for Machine Learning

Linear algebra, vector spaces, matrix transformations, multivariable calculus, and optimization.

View credential
Beyond delivery

Interests

Topics that keep me curious and influence how I approach systems, people, and technical problem-solving.

Machine Learning
Recommendation Systems
Computational Programming
Applied Mathematics
Social Psychology
Physics
AI & Science Fiction
Music
Let’s work together

Have an AI problem worth solving?

I’m interested in production AI/ML challenges involving agentic workflows, intelligent retrieval, scalable inference, and dependable backend systems.

© 2026 Nikhil Shingadiya · Designed for clarity, performance, and accessibility.