Resume

I'M DEVANSH RAI

AI Engineer

“Keep learning. Keep building. The earning will follow.”

Scroll
About & Core Philosophy

Engineering with curiosity, discipline, and a focus on real-world utility.

I am an AI Engineer and Full-Stack Developer driven by a passion for creating reliable, purposeful software. My work bridges the frontiers of modern artificial intelligence—multi-agent workflows, RAG architectures, and local model inference—with production backend systems.

Through engagements at Gaddr, STEMbotix, and research at RV University, I have contributed to production backend microservices, investigated zero-downtime server migrations, and conducted intrusion detection benchmarks using temporal flow modeling.

I prioritize clarity, clean abstractions, and measurable impact over buzzwords. Every project is an opportunity to learn deeper, build stronger, and contribute to meaningful technology.

Bachelor of Technology (Honors) in Data Science

2024 – 2028

Chhattisgarh Swami Vivekanand Technical University (CSVTU), Bhilai

Autonomous AI Systems

Designing multi-agent frameworks where specialized agents collaborate, gather competitive intelligence, and generate verified executive insights.

Production Backends

Developing high-throughput REST APIs, implementing selective recap engines, and supporting zero-downtime database and server migrations.

Applied Data & Research

Extracting temporal features from benchmark datasets, benchmarking state-of-the-art LLMs, and performing partition-based SQL analytics.

Career & Research

Where I've contributed and built.

Verified production engineering, AI embedded systems development, and academic cybersecurity research.

Backend & Systems Engineering Intern

Production Backend
Gaddr· Remote·Industry Experience
Production Engagement

Contributed to core production backend services, feature releases, and infrastructure stability.

  • Engineered production backend microservices in Python and Node.js, delivering reliable, high-throughput REST API endpoints and data serialization pipelines.
  • Developed and shipped the selective recap summary engine, aggregating and generating structured platform activity highlights for end users.
  • Assisted senior infrastructure engineers in server migration workflows, validating environment configurations, database backups, and executing zero-downtime service cutovers.
  • Optimized query performance and documented API contracts to enhance service resilience and cross-team collaboration.
Core Technologies:PythonNode.jsREST APIsServer MigrationSQLMicroservices

AI/ML Engineering Intern

Edge AI & LLM Systems
STEMbotix Pvt. Ltd.· Gandhinagar, Gujarat · On-site·Industry Internship
July 2026

Researched foundation LLMs, built an AI virtual lab, and engineered RAG-grounded code generation.

  • Benchmarked and systematically evaluated leading foundation LLMs (OpenAI, Claude, Gemini, Qwen, Llama, Mistral) to identify optimal speed, cost, and accuracy thresholds for automated hardware synthesis.
  • Architected and engineered the AI-based ESP32 Virtual Lab, integrating interactive browser circuit simulation, Blockly visual programming, and natural language-to-Arduino code synthesis.
  • Engineered a high-performance FastAPI backend interfacing with Ollama for local LLM inference, employing prompt engineering to achieve sub-second circuit schematic generation.
  • Implemented RAG-based semantic retrieval using vector embeddings to eliminate pinout hallucinations and guarantee valid circuit wiring topologies.
Core Technologies:FastAPIOllamaLocal LLMsRAGBlocklyESP32 / ArduinoPython

AI Research Intern

Cybercrime & ML Research
RV University· Bangalore · Remote·Academic Research
June 2026 – July 2026

Collaborated in a research team developing intrusion detection systems evaluated on benchmark datasets.

  • Collaborated within a 5-member research cohort to build an AI-based Cybercrime and Intrusion Prediction System evaluated on the CICIDS2017 benchmark dataset.
  • Engineered temporal traffic flow features, session statistics, and anomaly indicators to detect multi-class attack patterns across complex network traces.
  • Evaluated supervised machine learning classifiers and NLP-inspired sequence models to identify malicious network payloads with high precision.
  • Co-authored comprehensive research publications, methodology documentations, and model evaluation comparative reports.
Core Technologies:PythonScikit-learnCICIDS2017Temporal FeaturesPandasNLP / ML
Technical Skills

Tools & Engineering Stack

Grouped systematically across artificial intelligence, systems engineering, databases, and full-stack tools.

AI & Machine Learning

Multi-Agent SystemsLarge Language Models (LLMs)Retrieval-Augmented Generation (RAG)Prompt EngineeringVector Search & FAISSLocal Inference (Ollama)Explainable AI (XAI)Feature EngineeringSupervised & Unsupervised LearningModel Evaluation & Metrics

Backend & Systems

FastAPIPythonRESTful API DesignNode.jsServer MigrationAPI Performance OptimizationMicroservices ArchitectureLinux / BashZero-Downtime Deployment

Data & Database Engineering

PostgreSQLMySQLAdvanced SQL (CTEs, Window Functions)PandasNumPyScikit-learnData ModelingData Cleaning & EDAMatplotlib & Seaborn

Frontend & Full-Stack

TypeScriptReactNext.js (App Router)Tailwind CSSFramer MotionBlockly IntegrationHTML5 / CSS3Responsive UI Architecture

Engineering Tools & DevOps

Git & GitHubDocker (Basics)VS CodeStreamlitPostmanCI/CD WorkflowsTechnical Documentation
Selected Projects

Featured Engineering Systems

Real implementations in multi-agent orchestration, cybersecurity benchmark evaluation, and local LLM hardware simulation.

All Repositories on GitHub
Multi-Agent AI

Anshora

Autonomous Multi-Agent Business Development & Intelligence Platform

An autonomous multi-agent enterprise platform designed for strategic business development, competitive intelligence, and automated company analysis. Coordinates specialized AI agents to autonomously extract data, evaluate financials, model competitors, and generate executive reports.

Engineering Architecture
Orchestrates specialized autonomous agents via multi-agent communication protocols
Performs deep competitive intelligence and automated market research synthesis
Integrates vector search and RAG for hallucination-resistant financial report generation
PythonMulti-Agent OrchestrationFastAPILLMsVector SearchFAISS
View Source Repository
Applied ML & Security

AI-Based Cyber Crime & Intrusion Prediction

Network Threat Classification on Benchmark CICIDS2017 Dataset

A high-precision predictive cybersecurity and network intrusion detection system evaluated against the benchmark CICIDS2017 dataset. Utilizes temporal feature engineering and supervised machine learning architectures to classify multi-class cyber attacks.

Engineering Architecture
Rigorous data preprocessing and temporal flow feature extraction on millions of network packets
Benchmark validation on CICIDS2017 intrusion vectors with detailed precision/recall scoring
Co-authored technical documentation and research architecture
PythonScikit-learnCICIDS2017PandasTemporal EngineeringNumPy
View Source Repository
Edge AI & IoT

AI-Based ESP32 Virtual Lab

Interactive Hardware Simulation with Local LLM & RAG Code Generation

A virtual embedded systems engineering lab combining interactive circuit simulation, Blockly visual programming, and prompt-driven Arduino C++ code generation. Powered by a local Ollama inference engine and semantic RAG retrieval.

Engineering Architecture
Natural language prompts synthesized into executable Arduino C++ code and circuit wiring
Local inference execution via Ollama and FastAPI for secure, offline-capable embedded prototyping
RAG-grounded component pinout retrieval to eliminate electronic wiring errors
FastAPIOllamaRAGBlocklyArduino C++PythonReact
View Source Repository
Predictive Analytics

Flood Risk Decision Support System

Predictive Geospatial Modeling for Environmental Risk Mitigation

A machine learning-driven decision support system designed to quantify hydrological vulnerability and forecast regional flood risk probabilities, providing actionable insights for disaster preparedness and municipal response planning.

Engineering Architecture
In-depth exploratory data analysis (EDA) and multi-variable geospatial feature engineering
Predictive modeling pipeline generating calibrated vulnerability risk scores
Interactive Streamlit analytical dashboard for real-time risk scenario simulation
PythonScikit-learnStreamlitPandasMatplotlibSeaborn
View Source Repository
Data Systems

Advanced SQL Business Intelligence Engine

High-Volume Transaction Analysis & Revenue Optimization

Enterprise SQL analytical engine evaluating 20,000+ e-commerce transaction records. Employs common table expressions (CTEs), window functions (RANK, DENSE_RANK, ROW_NUMBER), and partition-based aggregations to uncover revenue trends.

Engineering Architecture
Engineered modular analytical queries across customers, orders, products, and regional sales
Generated executive KPI metrics including monthly growth velocity and customer lifetime value
Optimized query execution plans and database schema normalization
MySQLAdvanced SQLCTEsWindow FunctionsData Modeling
View Source Repository
Verified Credentials

Certifications & Honors

Direct verification links to official Microsoft credentials, comprehensive bootcamps, and industry job simulations.

Official Associate Certification2026

Microsoft Certified: Azure AI Apps and Agents Developer Associate — AI-103

Microsoft

Applied Skills Credential2026

Microsoft Applied Skills Credential

Microsoft

Verified Course Credential2026

Python for Data Science and Machine Learning Bootcamp

Udemy

Industry Job Simulation2026

BCGX Data Science Job Simulation

Forage / Boston Consulting Group

Hackathon Recognition

National Finalist — Synapse Data Science & ML Hackathon

Scaler School of Technology · Ascent 2026 — Selected as national finalist among premier engineering cohorts across India, competing in predictive machine learning and end-to-end data pipeline challenges.

Initiate Contact

Let's build something meaningful together.

Open to worldwide opportunities across AI, Software Engineering, and Full-Stack Development. Feel free to reach out directly.

Send Direct Email