Engineering
intelligence that
holds up in production.
I'm a recent BS Computer Science graduate from LUMS, specializing in AI, MLOps and applied AI systems. I care about the gap between a model that scores well on a benchmark and a system that actually works — reliable, deployable, and honest about its limits.
My work spans the full lifecycle: from designing a council-based vision framework that earned a 4.0 on my two-semester senior project and a peer-reviewed publication, to standing up a fully air-gapped RAG system for a public-sector client that cut internal retrieval time by roughly tenfold.
Across five internships I've built data-validation pipelines, lead-scoring models, LLM-powered text generation, and system-level tooling in C#. I'm comfortable owning a problem end-to-end — experimentation, evaluation, packaging and deployment.
Focus areas
Five teams, one constant — shipping.
Biome / XFlow
- Built config-driven data-validation pipelines in Great Expectations to validate complex datasets and publish automated Data Docs reports.
- Implemented and registered custom GX expectation plugins tailored to client data, surfacing failed rows for triage.
- Developed end-to-end runner scripts to load data, parse JSON configs, run ad-hoc validation, and build summary reports.
Excellence Delivered (ExD)
- Led design and development of a lead-generation prototype alongside the company's core team.
- Built a pipeline to ingest, clean, score and label incoming leads as Hot, Warm or Cold from predictive insights.
- Integrated an LLM-powered system to generate and send personalized outreach emails tailored to each lead.
LMKR
- Worked on LLM-based text generation, producing coherent, contextually relevant outputs for product use cases.
- Fine-tuned pre-trained transformer models via transfer learning to improve generation quality.
- Owned the model workflow end-to-end: training, evaluation, packaging and deployment.
NESCOM
- Managed Windows system-file operations in C#, practicing core system-level programming.
- Engineered a TCP server–client framework for reliable, high-throughput data transfer to a central server.
- Built and deployed a Windows Forms app that automates device scans for missing software and reports to a backend.
Renzym
- Built foundational OOP skills in C++ and C under mentorship from the project manager and company founder.
- Contributed to user-facing software with a focus on correctness and usability.
Where it started.
BS Computer Science
Coursework: AI · Machine Learning · Generative AI · Deep Learning · NLP · Data Science · Operating Systems · Databases · Software Engineering · Theory of Automata
O-Levels & A-Levels
O-Level: 4A* · 5A
Certifications & memberships.
What I work with.
Languages
AI & Machine Learning
Web & Application
Tools & Infrastructure
See the work behind the numbers.
Five projects spanning vision, retrieval, NLP and full-stack engineering — plus two research publications.