3rd Place Winner @ Algoryc OpenHome AI Hackathon

Engineering Autonomous AI Agents & Real-Time Voice Systems.

Hi, I'm Umar. BSAI Graduate & AI Developer at Poshnee Tech. I architect LLM agent workflows (LangChain/LangGraph), Voice AI telephony systems (Twilio Voice API, Whisper ASR), AI automations, and production MLOps pipelines with full-stack Python & FastAPI backend integration.

Explore AI Projects Download Resume
Umar Faizan - AI Engineer

Umar Faizan

AI & Voice Systems Engineer

LangChain & LLM Agents Voice AI (Twilio & Whisper) MLOps & AI Automation

3rd Place AI Winner

Algoryc OpenHome Hackathon

Voice AI & Telephony

Twilio Voice & Whisper ASR

LLM & AI Automation

LangChain, LangGraph & RAG

Production MLOps

Docker, AWS & CI/CD Pipelines

Portfolio Systems

Featured AI, Voice & Automation Work

Production-ready conversational agents, Voice AI telephony systems, AI automations, and MLOps pipelines.

OpenHome SDK | Twilio Voice API

Voice AI Smart Home Safety System

🥉 3rd Place Winner @ Algoryc Hackathon
Hackathon Winner — Voice AI

Panic Emergency Protocol (OpenHome Ability)

Integrates OpenHome with the Twilio Voice API to offer two emergency distress pathways: a covert silent alarm for security threats (e.g., robbery/intrusion using conversational trigger phrases) and an interactive conversational guide with real-time safety checklists and custom helper loops for medical or fire crises.

Twilio Telephony Dispatch Call
Covert & Interactive LLM Paths
Python Twilio Voice API OpenHome SDK LLMs AI Automation
OpenHome Platform
GitHub Repository

AI Speech & Language Platform

Final Year Project
EchoVerse Showcase
Speech Processing & ASR

EchoVerse AI Platform

Full-stack AI-powered English language learning app combining interactive lessons with real-time conversational AI features. Integrates GPU-accelerated Whisper speech recognition (ASR), LLM dialogue generation loops, and human-like text-to-speech (TTS) playback to provide real-time structural pronunciation grading and dynamic vocabulary feedback.

Whisper STT → LLM → TTS Pipeline
Multilingual Dynamic Feedback
Python Whisper ASR LangChain TTS Engines FastAPI
View Code Video Demo
AWS EC2 Deployment | Docker

380K+ Records Insurance ML

Automated CI/CD Live
MLOps & Cloud DevOps

Vehicle Insurance Cross-Sell Pipeline

End-to-end production MLOps pipeline predicting customer cross-sell interest on 380K+ records. Features modular 6-stage architecture, MongoDB Atlas integration, AWS S3 model registry, and automated GitHub Actions deployment to AWS EC2 via ECR containers.

MongoDB Atlas & AWS S3 Registry
Dockerized AWS ECR & EC2 CI/CD
Python FastAPI Docker AWS (EC2/ECR/S3) GitHub Actions MongoDB
View Code
sql-data-chat-pro.streamlit.app

Conversational SQL Assistant

Live Demo Active
DataChat Pro Showcase
Generative AI & LLMs

DataChat Pro — NLP to SQL

Natural language to SQL database querying system. Allows users to query complex SQL databases in plain English and automatically generates interactive visualization charts via Streamlit.

Automated Chart Visualizations
Prompt-to-SQL Guardrails
Python LangChain OpenAI API Streamlit SQLite
Live Demo View Code
superleadpro.com

AI Lead Enrichment Extension

Chrome Web Store Live
SuperLead Pro Showcase
Full-Stack Data Mining

SuperLead Pro Extension

Commercial lead extraction tool and Chrome Extension (Manifest V3). Features automated website enrichment, contact extraction backend, and AI lead qualification filters.

Manifest V3 Side Panel Architecture
Automated Data Mining
Python JavaScript Chrome Extension API REST APIs
Live Site
resume-analyzer.streamlit.app

LLM ATS Evaluation Engine

ATS Scoring Active
NextGen Resume Analyzer Showcase
AI Parser Agent

NextGen Resume ATS Analyzer

ATS evaluation scanner powered by DeepSeek and LLMs. Analyzes PDF CVs against job descriptions, outputs match percentages, and generates structured optimization suggestions.

DeepSeek LLM Integration
PDF Document Parsing
DeepSeek Streamlit PDF Plumber Python
Live Demo View Code
Technical Mastery

Core AI, Automation & Engineering Stack

Proficiency across conversational AI, telephony systems, LangChain/LangGraph agent orchestration, AI automation, and production deployment.

Voice & Conversational AI

  • Twilio Voice API & Telephony Integration
  • Whisper ASR (Speech Recognition)
  • Real-time Audio WebSockets
  • TTS (ElevenLabs / Coqui Engines)
  • OpenHome Voice Ability SDK

LLMs & AI Automation

  • LangChain & LangGraph Agent Frameworks
  • AI Workflow Automation & Tool Calling
  • RAG Pipelines (ChromaDB Vector Store)
  • OpenAI / Groq / DeepSeek LLM APIs
  • Custom Prompt Engineering & Guardrails

MLOps & Infrastructure

  • Docker Containerization
  • AWS (S3, ECR, EC2 Deployments)
  • GitHub Actions CI/CD Retraining
  • DVC & MLflow Experiment Tracking
  • MongoDB Atlas & SQLite Databases

Full-Stack AI Integration

  • Advanced Python Programming
  • FastAPI & RESTful API Architectures
  • HTML5 / CSS3 / JavaScript Frontend
  • Streamlit Rapid App Interfaces
  • Git & GitHub Version Control
Umar Faizan - AI Developer
Engineering Profile

About Umar Faizan

I am a BSAI (Bachelor of Science in Artificial Intelligence) graduate from NUML Islamabad with CGPA 3.61/4.00, specializing in Voice AI systems, autonomous LLM agents (LangChain/LangGraph), AI automation, and enterprise MLOps pipelines.

Currently, I serve as an AI Developer Intern at Poshnee Tech, engineering custom LLM assistants, training platforms, and FastAPI backend integrations. Previously, I was a Data Science Intern at 10Pearls Pakistan (AQI forecasting with automated GitHub Actions CI/CD) and served as a Section Leader (Teaching Assistant) for Stanford University's Code in Place program.

Recently, I secured 3rd Place at the Algoryc OpenHome AI Hackathon by building the Panic Emergency Protocol, a Voice AI smart home responder utilizing the Twilio Voice API.

3.61/4.0

BSAI CGPA

🥉 3rd

Algoryc OpenHome Hackathon

100%

Production & Code Focus

Download Resume PDF
Career Journey

Work Experience & Education

A timeline of my software engineering roles, academic credentials, and hackathon wins.

Work Experience

April 2026 – Present

AI Developer Intern

Poshnee Tech | Islamabad, Pakistan
  • Engineered MVP AI Assistant using Python backends, custom prompt engineering, and REST APIs for real-time guidance.
  • Built MVP AI training platform integrating LLM APIs to generate realistic customer conversations for trainee evaluations.
  • Collaborated with Agile cross-functional teams to prototype and deploy core AI functionalities.
Dec 2025 – Feb 2026

Data Science Intern

10Pearls Pakistan | Remote
  • Developed Air Quality Index (AQI) forecasting system utilizing environmental telemetry streams (82% accuracy).
  • Implemented automated ML pipelines with GitHub Actions, Docker, and MongoDB, reducing deployment cycles by 70%.
Apr 2026 – Jun 2026

Section Leader (Teaching Assistant)

Stanford University (Code in Place) | Remote
  • Mentored international students in Python syntax, runtime debugging, and systematic problem solving.
  • Led weekly interactive programming labs and maintained top-tier student feedback metrics.

Education & Achievements

July 2026

🥉 3rd Place Winner — OpenHome AI Hackathon

Organized by Algoryc | In-Person

Built Panic Emergency Protocol, an open-source Voice AI smart home safety Ability integrating Twilio Voice API, covert alarms, and LLM emergency guidance loops.

2022 – Jan 2026

Bachelor of Science in Artificial Intelligence (BSAI)

National University of Modern Languages (NUML), Islamabad
CGPA: 3.61 / 4.00

Focused on Machine Learning, Deep Learning, Natural Language Processing, Voice AI Systems, and Software Architecture.

2018 – 2021

Intermediate (Pre-Medical)

Punjab College, Mailsi
Marks: 96%
2016 – 2018

Matriculation

Government High School, Mailsi
Marks: 92%
Credentials

Certifications & Continuous Learning

Verified professional milestones validating theoretical depth and practical AI engineering mastery.

Stanford University

Machine Learning Specialization

Verify Credential
DeepLearning.AI

Deep Learning Specialization

Verify Credential
Oracle University

OCI 2025 AI Foundations Associate

Verify Credential
Google Career Certificates

Google Soft Skills Program

Verify Credential
FAQ

Frequently Asked Questions

Key engineering details for recruiters and engineering hiring managers.

Are you open to full-time remote or on-site AI Engineer / ML Developer roles?

Yes! I am actively looking for full-time AI Developer, Machine Learning Engineer, or Voice AI roles—whether remote or on-site. I am ready to bring my Python backend, LangChain, Voice AI, and MLOps skills to a collaborative team.

What experience do you have with Voice AI & Telephony integrations?

I have built production Voice AI applications using the Twilio Voice API, Whisper ASR (Speech-to-Text), real-time WebSockets, and TTS engines. My hackathon project (Panic Emergency Protocol) won 3rd Place at the Algoryc OpenHome AI Hackathon by deploying covert and interactive telephony alert streams.

How do you approach AI Automation & LangChain Agent workflows?

I use LangChain and LangGraph to build stateful multi-agent workflows, tool calling, and RAG pipelines connected to vector databases (ChromaDB). This enables automated decision loops, document processing, and AI workflows.

What is your MLOps toolset for automated deployments?

I build modular ML pipelines using Python, FastAPI, and Docker. I use DVC for data versioning, MLflow for experiment tracking, MongoDB Atlas for storage, and AWS (S3, ECR, EC2) with GitHub Actions for automated CI/CD daily model retraining cycles.

Where are you located, and how do you work with remote engineering teams?

I am based in Islamabad, Pakistan (UTC+5). I communicate asynchronously and synchronously via GitHub, Slack, Zoom, and video updates, aligning seamlessly with global team sprints.

Get in Touch

Let's build intelligent AI systems together.

Have an open AI Developer position, Voice AI project, or MLOps engineering challenge? Drop me a message. I'll get back to you within 24 hours.

Send a Message

Message Sent!

I will get back to you within 24 hours.