Qasim Saeed

Academic

Verbatim text

Qasim Saeed
+923127603619, mgasim04337@gmail.com
~
Gll/3, Islamabad .
..
Linkedln: httes:[lwww.linkedin.com[inlgasim-saeed-246a7524al
Education
Bachelor of Science (Artificial lntellii:i:ence)
Major:
!Deep Reinforcement Learning, Natural Language Processing, Computer Vision, LLMs, Agentic Al.
Punjab College of Science, Gujrat Punjab
F.Sc (Physics, Chemistry, Biology, Mathematics)
Matriculation (Physics, Chemistry, Biology, Mathematics)
Projects
Final Project: Adaptive Communication Framework for Multi-Agent Reinforcement Learning [PyTorch, MARLib,
PettingZoo, TorchRL]
Ensuring consistent communication in multi-agent reinforcement learning (MARL) by developing a robust protocol
that enables seamless collaboration between agents t rained at different times.
Se.me.~te.r Projec~:
Video Captioning on Soccer Videos [GPT-2, VideoCLIP, GenAI]:
Generating commentary of a soccer match by taking visual clips as inputs.
Multilingual Detection of Persuasion Techniques in Memes [ANN, Pytorch, Transformers]:
SEMEVAL 2024 Task 4, involves using techniques like BERT for text embeddings and ResNet for visual embeddings.
Knowledge Graphs [Protege, OWL]:
Knowledge graphs for research paper dataset .
PROJECTWORKBASE(PWB} (Node js, Express, MySQL, HTML, CSS]:
A software tool used for project management, issue tracking, and software development.
AirXonix [Assembly Language]:
Coded a 2D game in Assemblv wit h the imolementation of Deoth-First Search for situation analvsis.
Work Experience
Head of Al at Vector Inc., Islamabad
Oct 2024-Present
Worked with multiple projects like Video generation, Voice Cloning, Detection from videos.
Al Intern at AIO, Islamabad
Worked with tools like Rasa, LLM Funct ion Calling, Voice Transcription.
Jan 2024-March 2024
Skills & Tools
Professional Skills
Problem-Solving & Critical Thinking, Research & Report Writing (Overleaf, LaTeX), Adaptability &
~ontinuous Learning.
Technical Skills
~ NN s, RNNs, GANs, VAE, RL, Adversarial RL, Decentralized MARL, OpenCV, YOLO, Faster R-CNN,
W-Net, Image Segmentation, Conditional GANs, VAE, Sketch-to-Image Synt hesis, LangChain, LLM
ntegration, Vector Indexing.
Achievements
Second Runner Up in Hackathon organized by Vision RD at National Science & Technology Park.
Won the DataQuest challenge at NaSCon'23.
Trainings I Certification
Deep Reinforcement Learning (DeepMind YouTube Playlist) - Self-paced learning
LangChain & LLMs for Al Aoolications - Hands-on implementat ion
Activities
Assist in Running AGI Lambda Server on Discord
Follow youtube channels (Veritasium, Faisal Warraich, 3Blue1Brown, Edan Mayer)
Interests
History, AGI.
FAST NUCES ISLAMABAD CAMPUS

AI enrichment