Ajas' CV
Ajas Bakran
Senior Engineering Lead | AI Agents & LLM Evaluation | AI Reliability | Python
Profile summary
Accomplished Senior Engineering Lead with 10 years of progressive experience in AI reliability, LLM evaluation, and test automation. Track record of establishing evaluation frameworks that reduce hallucinations by 90 percent and designing adaptive datasets that increase complex dialogue coverage by 50 percent. Dedicated to directing cross-functional engineering teams and building scalable automation frameworks that accelerate production delivery.
Career highlights
AI Reliability and Evaluation: Established the LLM evaluation framework to validate AI outputs prior to production release, achieving a 90 percent hallucination reduction and building automated CI/CD validation workflows.
Agentic Workflow Expansion: Directed the expansion of the Deal Management Agent for natural-language auction package management, driving six-figure media spend and delivering agentic targeting capabilities across workflows.
Dataset Engineering and Coverage: Designed adaptive evaluation datasets using prompt engineering, increasing coverage of complex dialogue scenarios by 50 percent and reducing dataset creation time by 60 percent.
Automation Framework Architecture: Architected automation frameworks from scratch, decreasing test execution time by 10x and overall testing time by 40 percent while attaining 80 percent code coverage across API repositories.
Key skills
Professional experience
Directs cross-functional engineering teams, product owners, and developers, guiding technical strategy, release planning, and production delivery. Manages the execution of complex AI initiatives, automation frameworks, and quality engineering strategies across distributed systems.
- Directed the expansion of the Deal Management Agent for natural-language auction package management, driving media spend and delivering agentic targeting capabilities across workflows.
- Established the LLM evaluation framework to validate AI outputs prior to production release, achieving a 90 percent hallucination reduction and building automated CI/CD validation workflows.
- Designed adaptive evaluation datasets using prompt engineering, increasing coverage of complex dialogue scenarios by 50 percent and reducing dataset creation time by 60 percent.
- Architected automation frameworks from scratch, decreasing overall testing time by 40 percent while attaining 80 percent code coverage across API repositories.
- Mentored 8 SDETs and developers across planning, technical strategy, and engineering initiatives while coordinating sprint-level planning and cross-team delivery.
Directed test automation initiatives, establishing technical direction for framework development and troubleshooting. Reviewed technical specifications, translating requirements into validation strategies while coordinating with developers and test architects to integrate automated workflows into software delivery.
- Designed and implemented automation scripts and Jenkins CI/CD pipelines to support software delivery.
- Resolved complex technical issues within automation frameworks using Java and Selenium.
Guided engineering collaboration and regression testing within Agile SAFe structures, coordinating technical deliverables across distributed teams and clients.
- Enhanced test-data automation frameworks, reducing manual test-data creation time by 40 percent.
- Developed and maintained Selenium-based automation frameworks supporting regression and post-release validation.
Education
Electronics & Communication Engineering | CGPA: 7.69