Muneeb Ali

Full-Stack Engineer | AI Specialist | Architecting Scalable Multimodal Systems

Based in: Lahore, PKCurrently: Teaching Assistant — Database Systems, Dr. Sarwan Altaf Abbasi, FCCUExperience: 2 years

CS graduate from FCCU specializing in AI engineering and full-stack development. Experienced in architecting microservices, fine-tuning LLMs, and building real-time multimodal AI pipelines. Proven track record of delivering complex, scalable systems like Oratoria AI and AI-powered recruitment tools, with deep expertise in Python, Next.js, and the LLM ecosystem.

Experience

Teaching Assistant — Database Systems

Dr. Sarwan Altaf Abbasi, FCCULahore, PKpart time2025 – Present

Supported 200+ students on SQL, relational design, and normalization; improved course materials and streamlined assignment workflows

SQLRelational DesignNormalization

Associate Software Engineer

Coders AlleyLahore, PKworkMay 2026

requests per minute throughput: 1,000requests per minute throughput
reduction in latency: 20%reduction in latency

Summary

Develop and maintain full-stack web applications for client projects, working across Next.js/React frontends and Python-based backends. Design and build AI pipelines integrating LLM APIs, embeddings, and third-party data sources to power product features for client deliverables

Results

Optimized AI pipelines and full-stack features to handle high-traffic loads of 1000 requests per minute.
Reduced system latency by 20% through backend and pipeline optimizations.

Next.jsReactPythonLLM APIsEmbeddings

Software Engineering Intern

Pak Elektron Limited (PEL)Lahore, PKinternshipMay 2025 – Jul 2025

Built and maintained internal web applications using ASP.NET, C#, and SQL Server in a large enterprise environment

ASP.NETC#SQL Server

Skills

Technical

Python

Git

HTML/CSS

FastAPI

React

OpenAI API

REST APIs

Next.js

GitHub

PyTorch

JavaScript

TypeScript

SQL

TailwindCSS

shadcn/ui

Django

WebSockets

OAuth 2.0

PostgreSQL

SQLAlchemy ORM

Docker

Transformers

Hugging Face

Whisper

Vector Embeddings

Pandas

NumPy

Jupyter

Colab

ASP.NET

Wav2Vec2

MediaPipe

Express.js

Node.js

OpenCV

scikit-learn

SERP API

SQL Server

Linux

Alembic

Agile

Framer Motion

Chrome Extension API

Azure

Projects

Oratoria AI

Lead Developerpersonal

Leading development of an AI-powered communication coach delivering real-time feedback for debate, interview, and presentation training. Architected 3-service microservices backend with FastAPI, PostgreSQL, SQLAlchemy ORM, Alembic migrations, RESTful APIs, and WebSocket protocol for low-latency bidirectional communication which is optimized for 100+ concurrent users. Integrated multimodal AI pipeline (Whisper STT, Wav2Vec2 emotion detection, MediaPipe gesture analysis, GPT-5/Mistral) and built Next.js 15 frontend with OAuth 2.0 (Google, GitHub, Facebook) and glassmorphism UI

Next.jsFastAPIPostgreSQLDockerWebSocketsLLMsSQLAlchemy ORMAlembicWhisper STTWav2Vec2MediaPipeGPT-5MistralOAuth 2.0

LLM Fine-Tuning for Debate Evaluation

Researcher/Developeracademic

Fine-tuned 12 open-source LLMs (Llama 3, Qwen 2.5, OpenHermes 2.5, WizardLM) achieving human-aligned debate scoring across content, structure, argumentation, and delivery. Developed iterative prompt optimization strategy reducing scoring deviation by 40%+ across 3–5 iterations. Built data pipeline processing 500+ hours of debate audio; benchmarked model robustness across prompt variations

PythonPyTorchTransformersHugging FaceLlama 3Qwen 2.5OpenHermes 2.5WizardLM

Job Board

Full-Stack Developerpersonal

Built a full-stack job board platform with dedicated candidate, recruiter, and admin portals using Next.js and FastAPI. Developed an AI-powered resume builder and resume-scoring engine using OpenAI and vector text embeddings to evaluate and rank candidate profiles. Implemented semantic job-matching via vector embeddings and integrated SERP API for automated job and company data aggregation

Next.jsFastAPIOpenAIVector EmbeddingsSERP APIHybrid Search

Education

Bachelor of Science in Computer Science

Forman Christian College and UniversityComputer Science2022 – 2026

Coursework

Mathematics

Recognition

Awards

Semi-Finalist — Allied Bank 4th FinTech Hackathon

Allied Bank2025

Competed nationally; advanced to semi-finals with a voice-enabled AI banking solution targeting non-smartphone users, including AI-driven savings optimization and personalized marketing features

Patents & publications

Can Lightweight LLMs Judge Debate?

research paper2025

Submitted — ACL Student Research Workshop 2025. Investigated whether sub-8B open-source LLMs can reliably evaluate debate quality; developed iterative prompt optimization achieving human-aligned scoring without GPT-scale models

What drives the work

Motivations: Building scalable AI-driven systems
Optimizing high-traffic backend architectures
Developing multimodal AI applications
Objectives: Full-Stack Development
AI Engineering
Python Backend
LLM Fine-Tuning
Microservices Architecture
Real-Time Systems
Next.js Ecosystem
Multimodal AI
Scalable Web Apps

Hire MuneebReach out about a role, a contract or a conversation.For recruiters

Muneeb's twin is AI, it can make mistakes.

More people on Kanvis

Browse more people Create your own Kanvis