Hello, I'm
Khaled Badran
AI Engineer & Software Researcher
I build AI systems with a solid software engineering base and proper ML ops lifecycle.
What I Do
I design and ship AI systems, ranging from bilingual RAG architectures and agents to data anonymization pipelines. I approach every project with a software engineering mindset, focusing on building production-ready, maintainable systems that meet regulatory standards. I also love the creative side of the work and am constantly brainstorming new ideas.
Where I've Worked
Jan 2025 - Present
- Managed the full model lifecycle with CI/CD pipelines for cloud and on-premises deployment.
- Deployed 5 bilingual RAG systems on Azure for different banking departments.
- Built an on-premises bilingual OCR & NER pipeline for automated document extraction.
- Designed a talk-to-your-data LLM tool enabling natural language queries on business data.
- Contributed to designing an on-premises central AI platform (hardware sizing, orchestration, MCP).
- Built a bilingual chatbot handling 50+ banking tasks with over 90% accuracy.
- Developed a service anonymizing 20+ types of sensitive data for privacy compliance.
- Mentored junior developers through their first production feature deployments.
- Delivered technical workshops across departments to drive company-wide AI adoption.
- Built a multi-skill Microsoft Copilot agent using Copilot Studio and Power Apps.
Academic Background
MSc Software Engineering
Concordia University · Montreal, Canada
Outstanding Evaluation, Dean's Honour List. Thesis: Using ChatGPT to Augment Software Engineering Chatbots Dataset.
SE for AI Program Trainee
Concordia University · Montreal, Canada
Relevant Courses: Software Engineering for LLMs, Technology and Society.
BSc Software Engineering
Concordia University · Montreal, Canada
GPA: 3.96, Dean's Honour List.
Research & Papers
Published at top-tier venues with over 480 total citations.
K. Badran et al. · MSc Thesis, Concordia University, 2024
Explores using ChatGPT to generate additional training data for the NLU component of software engineering chatbots. Retraining on the augmented datasets improved F1-scores by 3.9% to 11.6% across four SE datasets, with generated examples being syntactically diverse while preserving the original semantic meaning.
A. Abdellatif, K. Badran, E. Shihab · Empirical Software Engineering (EMSE), 2020
Presents MSRBot, a bot that lets developers query software repositories using natural language instead of requiring specialized mining expertise. Supports 15 common developer questions and was shown in a user study to simplify information extraction and reduce task completion time.
A. Abdellatif, K. Badran, D. Costa, E. Shihab · IEEE Transactions on Software Engineering (TSE), 2021
Evaluates four NLU platforms (IBM Watson, Dialogflow, Rasa, Microsoft LUIS) on intent classification, confidence scoring, and entity extraction for SE chatbots. IBM Watson achieved the best F1 above 84%, while Rasa provided the most reliable confidence scores.
A. Abdellatif, D. Costa, K. Badran, R. Abdalkareem, E. Shihab · MSR 2020
Analyzes Stack Overflow posts to map the challenges chatbot developers face. Identifies 12 discussion topics across five categories, finding that development, integration, and NLU training are the most common pain points and that the chatbot community still lacks maturity compared to established SE fields.
K. Badran, P-O. Cote, A. Kolopanis, R. Bouchoucha, A. Collante, D. Costa, E. Shihab, F. Khomh · IEEE Computer, 2023
Investigates whether combining multiple fairness pre-processing algorithms into an ensemble can produce a more robust approach to mitigating ML bias. Finds that different algorithms employ conflicting fairness concepts, creating trade-offs between fairness and accuracy, and offers practical guidance for practitioners.
Things I've Built
Arabic Music Generation
Fine-tuned ACE 1.5 using LoRA on Arabic music to produce authentic melodies, rhythms, and instrumentation. Built a lightweight training and inference workflow for rapid iteration across genres.
Technical Toolkit
Languages
Platforms & Tools
Expertise
Honours & Scholarships
Fonds de Recherche du Quebec Graduate Scholarship
2023
NSERC Canada Graduate Scholarship
2022
Top Concordian Graduate Scholarship
2022
Training & Workshops Delivered
AI-Assisted Thesis Writing
Graduate-level workshop on using LLMs for thesis writing, covering effective prompting techniques and ethical considerations.
Improving Your Daily Workflows Using AI
Workshop for KFH staff on practical LLM applications in daily work, including Microsoft Copilot within the Microsoft suite.
Community & Service
Workshop Organizer
Women In Engineering Conference
Designed a workshop for high-school students introducing ML concepts. Created 16 chatbots emulating famous people to gamify the experience.
Web Chair
MSR 2023 (Mining Software Repositories)
Managed the conference website.
Web Chair
BotSE 2023 (Bots In Software Engineering)
Managed the workshop website.
Let's Connect
Have an idea or want to collaborate? Reach out.