Who is Arshad Shaikh?

Arshad Shaikh (Arshad Shafik Shaikh) is an Indian AI/ML Engineer, third-year B.Tech student in Artificial Intelligence & Machine Learning at Zeal College of Engineering and Research (SPPU, Pune), and the Founder of Taskas, a digital solutions and AI platform serving 50+ clients with ₹4L+ in revenue.

What is Taskas founded by Arshad Shaikh?

Taskas is a digital solutions and AI automation platform founded and engineered by Arshad Shaikh in Pune, Maharashtra. It has delivered software and web infrastructure to 50+ clients with 6 strategic partnerships and ₹4L+ in revenue.

What is Arshad Shaikh's educational background and CGPA?

Arshad Shaikh is pursuing his B.Tech in Artificial Intelligence & Machine Learning (2024–2028) at Zeal College of Engineering and Research, affiliated with Savitribai Phule Pune University (SPPU), holding an 8.45 / 10 CGPA and serving as Class Representative.

What are Arshad Shaikh's primary technical skills and projects?

Arshad Shaikh specializes in Python, PyTorch, YOLOv8 computer vision pipelines, FastAPI backend services, LLM multi-tool agent architectures (NOOR, NEXUS AI), Three.js WebGL graphics (3D AI Classroom), and has been selected twice for the national Smart India Hackathon (SIH 2025 & SIH 2026).

Arshad Shaikh — AI & ML Engineer

Engineering Intelligent Systems & Shipping Real Products.

I'm Arshad Shafik Shaikh — a third-year B.Tech AI & ML student at Zeal College of Engineering & Research (SPPU, Pune), Founder of Taskas, and an AI Engineer focused on shipping reliable machine learning, computer vision, and backend systems.

From training custom YOLOv8 models during my software engineering internship at Sun Fibo Technology, to architecting modular LLM backends with FastAPI and competing at the Smart India Hackathon (2025 & 2026), I combine rigorous algorithmic understanding with practical product ownership.

ACADEMICS

B.Tech AI & ML • CGPA 8.45

STARTUP

50+ Clients • ₹4L+ Revenue

INTERNSHIP

Sun Fibo Tech (AI/CV)

HACKATHONS

2x Smart India Hackathon

CORE STACK

Python • PyTorch • FastAPI

AVAILABILITY

Open to AI/ML Internships

My
Expertise

I design and build digital experiences where design, code, and motion work as one. From expressive interfaces to high-performance AI systems.

Drag / Hover 3D Scene

Python AI & ML Three.js React.js GSAP Flask Tailwind

WORK

HTML5 CSS3 JavaScript React.JS Python GSAP Three.js Flask

UI/UX INTERACTIVE WEB CREATIVE DEVELOPMENT MOTION DESIGN AI ARCHITECTURE

AI, SPEECH RECOGNITION & 3D ROBOTICS — 01

NEXUS AI Virtual Assistant

An intelligent voice-controlled assistant paired with an interactive 3D humanoid robotics interface. Built to handle continuous speech recognition, multi-intent NLP dispatching, and hands-free desktop automation.

Problem: Standard voice interfaces lack spatial presence and suffer from latency bottlenecks during continuous streaming audio capture.

Approach: Engineered a multi-threaded Python audio processing pipeline with SpeechRecognition and NLP intent matching, linked to a WebGL 3D robot that responds dynamically with spatial head/gaze tracking.

What I Built: Full speech audio capture thread, NLP dispatcher, automated OS workflows, and interactive 3D WebGL robot canvas.

<800ms Intent Dispatch Real-time Audio Stream 60 FPS 3D WebGL

Python SpeechRecognition NLP Spline 3D Multi-Threading

Explore Project ↗ Code ↗

3D AI Robot Interface

AI BACKEND & LLM ORCHESTRATION — 02

NOOR — AI Assistant Backend

A modular, production-grade AI assistant backend connecting large language models to external tool execution schemas, context-aware memory, and persistent relational storage.

Problem: Basic LLM wrappers fail at stateful multi-step tasks, tool reliability, and context drift over long conversations.

Approach: Engineered a modular FastAPI service layer with Pydantic validation, structured function calling, asynchronous background job handling, and dual SQLite/PostgreSQL persistence.

What I Built: Complete backend architecture, tool dispatcher, session context manager, and asynchronous API endpoints.

Multi-Tool Calling Stateful Memory Async Task Queue

Python FastAPI LLM APIs PostgreSQL Pydantic

Explore Project ↗ Code ↗

COMPUTER VISION & DEEP LEARNING — 03

YOLOv8 Real-Time Person Detection

Production computer vision pipeline developed during my software engineering internship at Sun Fibo Technology. Features custom-annotated dataset curation, hyperparameter optimization, and low-latency edge inference.

Problem: Detecting human subjects in cluttered indoor feeds with occlusions and strict real-time throughput constraints.

Approach: Annotated custom training datasets in Roboflow, applied mosaic and HSV color augmentations, trained custom YOLOv8 weights over 100 epochs, and performed error analysis on false negatives.

What I Built: Data pipeline, model training scripts, confusion matrix validation, and real-time OpenCV edge inference loop.

30+ FPS Edge Inference High mAP@0.5 Validation Industry Tested

YOLOv8 PyTorch Roboflow OpenCV Python

Internship Project ↗ GitHub ↗

SPATIAL COMPUTING & INTERACTIVE 3D — 04

3D AI Classroom (Three.js)

Interactive 3D educational environment showcasing technical range in browser graphics. Features an animated AI teacher avatar (`Emilian`), Web Speech API voice I/O, dynamic chalkboard rendering, and offline knowledge fallback.

Problem: Remote learning lack spatial immersion and interactive tutor responsiveness.

Approach: Built a full Three.js/R3F virtual classroom with GLTF skeletal animation mixers, dynamic lighting, OrbitControls exploration, and dual OpenAI/Gemini voice generation.

What I Built: Complete 3D scene architecture, skeletal animation controller, camera management, and multimodal voice integration.

3D Spatial Environment Voice Speech I/O Offline Fallback

Three.js React Three Fiber Gemini API WebGL Speech API

Explore Project ↗ Code ↗

3D Classroom Environment

FOUNDER CASE STUDY & PRODUCT OWNERSHIP — 05

TASKAS Digital Solutions

Full-stack digital solutions and AI platform founded and scaled by Arshad Shaikh. Features an automated project cost estimator, lead capture pipeline, and client onboarding funnels delivering production software.

Problem: Small businesses and creators need digital presence and automated software without bloated agency overhead.

Approach: Engineered a streamlined web application with interactive cost estimation, automated Google Sheets & email lead dispatch, and scalable cloud hosting.

What I Built: Full-stack product, customer intake algorithms, UI design, client CRM integrations, and business development.

50+ Clients Served 6 Partnerships ₹4L+ Revenue

Next.js / JS AI Cost Estimator Lead Automation Tailwind

Live Platform ↗ Code ↗

NATIONAL HACKATHON INNOVATION — 06

SIH AI Systems & Mobile Architecture

Two consecutive national selections for the Smart India Hackathon: an AI-based early warning student dropout prediction model (SIH 2025) and an offline-first commuter travel-data mobile architecture (SIH 2026).

Problem: Predicting academic dropouts before occurrence, and collecting uninterrupted commuter transport data in low-connectivity zones.

Approach: Trained supervised classification models on structured socio-demographic datasets; architected an offline-first mobile app with SQLite local caching and resilient background sync.

What I Built: ML feature engineering, model evaluation metrics, mobile system architecture, and technical pitch presentations.

2x National Selection Offline-First Sync Predictive Modeling

Scikit-Learn Pandas React Native SQLite Python

Explore SIH Solutions ↗ GitHub ↗

Arshad Shaikh — AI/ML Engineer | AI / ML

ENGINEER

Engineering Intelligent Systems & Shipping Real Products.

B.Tech in Artificial Intelligence & Machine Learning

Smart India Hackathon (SIH) 2026

Smart India Hackathon (SIH) 2025

Google Skills & Professional Certifications

Founder & Technical Lead

Software Engineer Intern (AI & Computer Vision)

Content Writer Intern

Looking for an AI/ML Intern or Junior Engineer Who Can Ship?

My
Expertise

Creative Development

Motion & Interaction

AI & Intelligent Systems

Modern Web Apps & APIs

WORK