Kaan - Resume

Kaan Un

AI Engineer and Forward Deployed Engineer

I build production AI systems with agent orchestration, evaluation loops, and operational guardrails. Founder of Traction IT, where the product is a multi-agent GTM stack. Prior work spans forward deployed engineering, MLOps, and Kubernetes-first platform rebuilds for enterprise banking clients.

Experience

Founder & GTM Engineer - Traction IT

2025 - Present - Toronto, ON

  • Engineered enrichment and personalization pipelines in Clay and Apollo, wired to multi-agent workflows in Cursor for account research and opener generation at scale.
  • Built a four stage system covering buyer signal mining, qualification, founder voice outreach, and retention triggers that catch churn before it appears.
  • Stood up custom outbound over SMS and phone, and shipped the marketing site on Next.js and Vercel.

Forward Deployed Software Engineer - Masraff International

2024 Jul - 2025 Mar - New York, Remote

  • Wore multiple technical hats across frontend development, data engineering, and AI/OCR R&D while maintaining consistent stakeholder communication.
  • Fine tuned LLMs for OCR and benchmarked a custom Llama model, accelerating project delivery by 60%.
  • Orchestrated 30+ PySpark data flows with AWS Step Functions, and embedded MLflow tracking for every model scoring step to give full lineage from raw data to predictions.
  • Trained under Ex-Palantir leadership on HyperAuto methodologies, cutting constraint identification and action planning from weeks to under an hour.

Cloud DevOps Engineer - Netoloji Software International

2023 Feb - 2024 Jun - Remote

  • Re-platformed a legacy SQL Server monolith into a Kubernetes-first serverless architecture with Helm-driven GitOps pipelines, giving ~200 banking users 99.9% availability.
  • Built a reliability and disaster recovery layer with RPO under 10 minutes, quarterly game-day restores, TLS-only ingress, and KMS-encrypted data paths via scoped IAM roles.

MLOps Engineer - Cogent Integrated Business Solutions

2023 Mar - 2023 Oct - MI, United States

  • Automated a fully serverless scraping and scoring pipeline chaining AWS Lambda functions with a Random Forest scorer, cutting researcher effort by 80%.
  • Baked CI/CD and experiment loops into Lambda, publishing the model as a Lambda Layer and rerouting traffic for A/B label-quality tests across 1M+ rows.
  • Directed rollout and operations for 15 researchers and data scientists, compressing idea-to-production lead time by 60%.

Core Organizer - AI Tinkerers Ottawa

2024 - 2026 - Ottawa, ON

  • Built and demoed agent-based systems using Claude, Gemini, and Sonar at community workshops, with trace instrumentation and monitoring.
  • Presented platform-agnostic agent skill architecture operating across Cursor, Claude Code, and Antigravity.

Education

Honours BSc. Applied Mathematics - University of Ottawa

BSc. Computer Science (Data Science) - University of Ottawa

Certifications

  • Cursor Ambassador
  • Clay Ambassador
  • Lovable Ambassador
  • AWS Cloud Practitioner

Links

linkedingithubwebsitetwittertractioncal