Looking for a job

mirza anas

@mirzaanas

Senior AI/ML Engineer at Sybrotix, delivering production RAG systems for international clients with four integrated LLM providers.

Canada

Message

At Sybrotix, I architected and deployed a production AI system for two international clients. It connects document ingestion and preprocessing with Python/FastAPI services, vector retrieval, LLM inference, validation, and user-facing reporting.

I built RAG pipelines for grounded question answering, source references, and structured summaries, and integrated four LLM providers through reusable orchestration services. I also fine-tuned a Llama model for domain-specific tasks and evaluated its grounding, relevance, and response validity.

At École de technologie supérieure (ÉTS), Université du Québec, I train and benchmark deep-learning models on UAV flight datasets as a PhD candidate. My reproducible research pipelines cover preprocessing, feature engineering, model comparisons, and evaluation.

Earlier, at Pakistan Aeronautical Complex (PAC), I worked on Python/C++ applications, real-time telemetry, machine-learning prototypes, and operational analytics for aerospace systems. I also contributed to peer-reviewed research at COMSATS University Islamabad on machine-learning approaches to aircraft fuel-consumption prediction.

Experience

Work history, roles, and key accomplishments

Current

Senior AI/ML Engineer

Current

Sybrotix

Apr 2026 - Present (7 months)

Architected and deployed a 6-layer production AI system for 2 international clients, integrating 4 LLM providers and fine-tuning Llama for domain-specific tasks. Owned end-to-end delivery including Docker-based deployment, CI/CD, and client-facing release verification.

Python fastAPI RAG LLM Vector Search Docker Kubernetes CI CD Git MLFlow Prompt Engineering Llama Fine Tuning Tool calling Schema Validation Retries Fallbacks Request Tracing Monitoring Cost monitoring

Current

Machine Learning and Autonomy Research Scientist

Current

École de technologie supérieure (ÉTS), Université du Québec

May 2023 - Present (3 years 6 months)

Trained and benchmarked 4 deep-learning model families on UAV flight datasets, publishing MAE of 3.38 W / 7.52 W and R of 0.9931 / 0.9668. Built reproducible model-development pipelines and researched multi-agent decision systems and model inference trade-offs.

Python PyTorch scikit learn Deep Learning LSTM GRU Transformers Hybrid ONNX Runtime TensorRT CUDA NVIDIA Jetson Reproducible Experiments Mentorship

Senior Design Officer

Pakistan Aeronautical Complex (PAC)

Apr 2017 - Apr 2023 (6 years)

Delivered 6 years of engineering work on Python/C++ applications, real-time telemetry, software integration, machine-learning prototypes, and operational analytics for aerospace systems. Supported 5 verification stages and mentored engineers across software, AI, integration, and testing.

Python C++Machine Learning Computer Vision Anomaly Detection Telemetry Software Integration Verification Mentorship Code Review

AI Research Engineer

COMSATS University Islamabad

Jan 2019 - May 2021 (2 years 4 months)

Compared 4 ML approaches across 3 flight phases for aircraft fuel-consumption prediction, using repeatable datasets and comparative performance assessment. Evaluated models using MAE, RMSE, and R and contributed to peer-reviewed research.

Machine Learning Python Feature Engineering Model Evaluation MAE RMSE R Research

Education

Degrees, certifications, and relevant coursework

École de technologie supérieure (ÉTS), Université du Québec

PhD Candidate, Engineering (Applied Research)

2023 -

PhD Candidate in Engineering (Applied Research) with final defense pending. Research focuses on deep learning, autonomous systems, and multi-agent AI.