Applied AI engineering lab based in Multan, Pakistan.
We train engineers to build and deploy real AI systems — not theory, not tutorials, not certificates.
- End-to-end AI applications
- LLM and RAG pipelines
- Speech AI systems
- NLP research tools
- MLOps infrastructure
-
Low Resource NLP
-
Data-Centric Roman Urdu NLP: High-Quality Dataset Curation, Privacy-Preserving Embeddings, and State-of-the-Art Model Benchmarking. Read here
-
RUDaSA: Roman Urdu Dataset for Sentiment Analysis — A Large-Scale, Curated Corpus with Privacy-Preserving Embeddings and Competitive Benchmarking of Transformer Models. Read here
-
RUEmoCorp: A Large-Scale Roman Urdu Emotion Corpus with Cross-Institute Annotation Validation and State-of-the-Art Emotion Classification. Read here
-
-
Speech AI
-
Modeling Vocal Fatigue as Embedding-Space Deviation Using Contrastively Trained ECAPA-TDNNs. Click here
-
Continuous Vocal Load Monitoring in Professional Voice Users: Development and Occupational Validation of an Automated Assessment System. (Under Review at Journal of Voice)
-
-
Human Factors and Ergonomics
- Ergonomic Interventions and Cognitive Workload in Healthcare Settings: A Structured Qualitative Case Study Using Cognitive Systems Engineering (Under Review at Journal of Applied Ergonomics, Elsevier)
-
RUEmoCorp (Largest curated Roman Urdu's Emotion Corpus) HuggingFace | Harvard Dataverse
-
Roman Urdu Sentiment Corpus (Largest curated Roman Urdu Sentiment Corpus) HuggingFace | Harvard Dataverse
- ECAPA-TDNN-VHE (Vocal Health Encoder with 2.5x performance over baseline ECAPA-TDNN) HuggingFace
- Roman Urdu Emotion Classifier (Current State of the Art) HuggingFace
- Roman Urdu Sentiment Classifier (Current State of the Art) HuggingFace
Muhammad Khubaib Ahmad — AI Research Engineer
Building the next generation of AI/ML engineers from Pakistan.
