Data Engineer | AI Systems Builder | Distributed Systems Enthusiast

I build scalable data infrastructure for AI systems.
From ingestion pipelines to distributed computing and GPU-accelerated workloads, I focus on designing high-performance architectures that power intelligent applications.

Currently pursuing a BS in Artificial Intelligence (FAST-NUCES) with a strong emphasis on:

• Data Engineering
• Distributed Systems
• AI Infrastructure
• High-Performance Computing

• Designing scalable ETL pipelines for AI workloads
• Building distributed and microservice-based AI systems
• Exploring GPU-accelerated data processing (CUDA, RAPIDS, parallel computing)
• Studying large-scale RAG systems and multi-agent reasoning
• Strengthening cloud-native architecture skills (AWS, Azure, GCP)

• ETL pipeline design & optimization
• Data scraping & transformation
• Data curation for ML pipelines
• Data governance & quality assurance

• Microservices architecture
• Containerization (Docker, Kubernetes)
• CI/CD pipelines
• Cloud-native deployments

• NLP systems (BERT variants, embeddings, RAG)
• PyTorch, TensorFlow, Scikit-learn
• Multi-agent reasoning frameworks
• GPU-based acceleration & parallel computing

• C#, .NET, Java
• Python (FastAPI, Flask)
• Node.js, Express
• REST & gRPC APIs

Microservice-based NLP system that generates creative stories from a topic and converts them into narrated audio.

Tech Stack:
Python, FastAPI, gRPC, Databricks Dolly v2, edge-TTS, Gradio, Docker

Flask-based demonstration of quantum-resistant encryption using modern KEM algorithms.

Tech Stack:
Python, Flask, liboqs-python
Algorithms: ML-KEM-512, FrodoKEM, BIKE, McEliece

Designing and evaluating a retrieval-grounded multi-agent debate framework to analyze:

• Knowledge drift
• Self-consistency
• Evidence alignment
• Convergence behavior

Comparative evaluation against:

• Single-Agent RAG
• Self-Consistency
• Tree-of-Thoughts

Focus: Stability vs. drift under noisy retrieval pipelines.

• Optimized ETL pipelines for AI model training
• Led interns and distributed data engineering tasks
• Built structured failure management systems
• Delivered insights from large-scale scraped datasets

• Worked on Urdu Word Sense Disambiguation
• Experimented with mBERT, mDistilBERT, RobertaUrdu
• Contributed research for academic publication

Languages:
C#, Java, Python, C++, JavaScript, TypeScript

Data & ML:
PyTorch, TensorFlow, Scikit-learn, NLTK, SpaCy, NumPy, Pandas

Cloud & DevOps:
Azure, Docker, Kubernetes, CI/CD

Databases:
CosmosDB, MSSQL, NoSQL

Systems:
Unix, Shell, System Administration

• Architect large-scale AI data platforms
• Contribute to open-source distributed systems
• Build GPU-optimized AI infrastructure
• Advance real-time streaming & serverless architectures
• Develop reliable, scalable AI systems beyond experimentation

• 🔗 LinkedIn: https://www.linkedin.com/in/talha-syed-b49304219/
• ✍️ Blog: https://substack.com/@bitwisetitan
• 💻 GitHub: https://github.com/BitwiseTitan
• 📧 Email: talha.syed1@outlook.com

Data is not just fuel for AI.
It is the architecture of intelligence.