Ubaith
Sherif
Focus:AI Engineer & Agentic Systems
Studying AI & Data Science in Coimbatore. Building multi-agent workflows, retrieval platforms, and production Next.js applications focused on latency, state, and reliability.
01b — About
I deconstruct systems to understand how they actually work, then rebuild them into things that are useful.
How I think
Data flow over isolated functions. Edge cases over happy paths. Failures are just lessons I haven't documented yet.
What I build
End-to-end AI systems — multi-agent workflows, retrieval pipelines, computer vision, and full-stack products around them.
Where I'm heading
From someone who solves isolated problems to an engineer who can architect, implement, and deploy complete AI products.
Outside the terminal
Football, competitive gaming, and visual design. Same instinct every time: read the pattern, make the call.
“Engineering at the intersection of AI models, clean architecture, and real-world products.”
Selected Projects
Other Systems
Decision Intelligence
InsightAI Enterprise
Multi-Agent Decision Intelligence Platform with real-time analytics, semantic search, and AI-powered business insights.
Workflow Automation
Multi-Agent Business Automation
Human-reviewed automation for sales, support, and reporting with approval gates and full auditability.
Research Automation
AI Company Research Assistant
AI-powered deep research on companies using web search, site crawling, PDF generation, and Discord integration.
NLP / ML
Email Support Ticket Categorizer
NLP-based support ticket categorization using TF-IDF, Linear SVM, confidence scoring, and priority tagging.
Natural Language Analytics
InsightAI Analytics Agent
Ask questions over business data and review the query, chart, and explanation before acting.
Meta AI / Code Generation
Self-Writing AI Engineering Platform
AI engineering platform that automates code generation, testing, and documentation workflows.
03 — Founder Experience
Work Experience
Unscript Labs
AI-Focused Technology Studio
01 / Overview
Co-founded Unscript Labs to build AI-powered products and digital solutions for real business problems — spanning web development, LLMs, RAG pipelines, multi-agent workflows, computer vision, and automation.
Full Lifecycle Ownership
Leading conceptual design, scalable architecture, AI orchestration, and production deployment across client and internal products.
02 / What I Do
Featured Work
What We Build
Websites
Web Design & Development
// PRODUCTION_WEBSITEconst site = {stack: 'Next.js + Tailwind',deploy: 'Vercel / CF Workers',status: 'LIVE',};
Designing and developing modern, responsive, production-ready websites and digital experiences for businesses, products, and organizations.
AI Chatbots
Conversational AI Systems
// CHATBOT_PIPELINEconst bot = {retrieval: 'Hybrid RAG',memory: 'MemorySaver',status: 'DEPLOYED',};
Building intelligent conversational AI systems using LLMs, RAG, knowledge bases, APIs, and custom workflows to create context-aware user experiences.
AI Software
Custom AI-Powered Applications
// AI_SOFTWARE_SYSTEMconst product = {ai: 'LLM + Vision',backend: 'FastAPI',status: 'PRODUCTION',};
Developing custom AI-powered software and automation solutions that solve specific business and operational problems with precision.
“Don’t just build what technology can do. Build what people actually need.”
— Ubaith Sherif, Co-Founder · Unscript Labs
Internships
Artificial Intelligence Intern
Web Epic Technologies Private Limited
June 2025 – July 2025
India
Worked on AI and machine learning application development with a focus on the full ML lifecycle from data collection through model evaluation.
Green Skills AI Internship
AICTE – Shell & Edunet Foundation
October 2025 – November 2025
India (Remote)
AI-focused internship through AICTE's national Green Skills program (supported by Shell India), building a data-driven ML project addressing real-world sustainability challenges.
Technical Stack
Technologies I work with daily across the full AI-to-product pipeline.
Academic Background
2022 - 2026
Bachelor of Technology
Artificial Intelligence and Data Science
United Institute of Technology
Coimbatore, India
Focus Areas
Academic foundation in artificial intelligence, data science, machine learning, backend engineering, and software development.
Coursework
AI-Enabled Intelligent Teacher Robot for Automated Attendance, Personalized Learning Assistance, and Real-Time Knowledge Retrieval
Automated Attendance · Personalized Learning Assistance · Real-Time Knowledge Retrieval
Project Overview
An AI-enabled intelligent teacher robot designed to automate classroom attendance, provide personalized learning assistance, and retrieve real-time knowledge through an AI-powered retrieval system. The system combines computer vision, voice interaction, large language models, and Retrieval-Augmented Generation to create an interactive educational assistant for students and instructors.
Core Capabilities
Technical Architecture
Computer Vision
InsightFace · ArcFace · SCRFD · OpenCV
AI / LLM / RAG
OpenAI APIs · Google Gemini · ChromaDB
Voice
Deepgram · Whisper (STT / TTS)
Backend
FastAPI · Node.js · PostgreSQL · SQLite
Frontend
React · Vite · Tailwind CSS
Edge Hardware
Raspberry Pi Integration
Research Publication
AI-Enabled Intelligent Teacher Robot for Automated Attendance, Personalized Learning Assistance, and Real-Time Knowledge Retrieval
International Journal for Research in Applied Science & Engineering Technology (IJRASET) · Published
Presents the design, engineering architecture, and experimental validation of an AI-Enabled Intelligent Teacher Robot. Details the computer vision pipeline for automated face recognition attendance, RAG-backed contextual knowledge retrieval, and voice-driven educational assistance.
Published Writing
The 5% Problem: Why Agent Failure Rate Isn't the Metric That Matters
Ubaith Sherif · AI Agents / System Reliability
Examines why raw failure rates are insufficient for evaluating autonomous AI agents, highlighting error impact, compounding failure modes, and resilient agent architecture.
Context Engineering: Replacing Prompt Engineering with a New Architecture
Ubaith Sherif · Applied GenAI / System Architecture
Explores the architectural transition from prompt tweaking to structured context engineering, retrieval orchestration, and stateful agent coordination in production AI systems.
Credentials & Training
Claude 101
Anthropic
Introduction to Claude Cowork
Anthropic Education — Anthropic
What Is Generative AI?
LinkedIn Learning
Official Practice Question Set: AWS Certified Generative AI Developer – Professional (AIP-C01)
AWS Training & Certification
Artificial Intelligence & Data Analytics Internship
Edunet Foundation
British Airways – Data Science Job Simulation
Forage
IBM SkillsBuild Data Analytics Certificate
IBM
Data Science & Analytics
HP LIFE
Quantium – Data Analytics Job Simulation
Forage
Common Questions
Direct insights on AI engineering, system architecture, technical background, and project collaboration.
I specialize in AI Engineering, Generative AI, Retrieval-Augmented Generation (RAG), agentic systems, FastAPI backend architectures, and production full-stack web development. My focus is on delivering evaluated, reliable, and auditable software rather than prototype demonstrations.
I build AI-native applications, multi-agent orchestration workflows with LangGraph, hybrid RAG pipelines, LLM-powered enterprise platforms, computer vision applications (face recognition, edge detection), and intelligent workflow automation platforms with human-in-the-loop validation.
My primary toolkit spans Python, FastAPI, TypeScript, Next.js, React, LangGraph, LangChain, Qdrant, ChromaDB, PostgreSQL (pgvector), MongoDB, Redis, Docker, and foundation model APIs (OpenAI, Anthropic Claude, Google Gemini).
It is my capstone engineering project and published research paper (IJRASET). The system unifies automated face recognition attendance (InsightFace), personalized student learning assistance, real-time knowledge retrieval via RAG, and natural voice interaction (Deepgram/Whisper) backed by FastAPI and Raspberry Pi edge hardware.
You can explore the featured engineering systems and case studies directly in this portfolio's Engineering Works section, and access source code, repositories, and technical contributions on GitHub at github.com/Ubaith444.
Get In Touch
Open for Collaboration
Seeking AI engineering roles, multi-agent system architecture, or technical advisory for high-impact software products.
Direct Reachout