A
82

Applied AI Researcher / Founding Engineer

6y relevant experience

Qualified
For hiring agencies & HR teams

EU engineers, ready to place with your US clients

Pre-screened on AI. Remote B2B contracts. View 5 full profiles free — AI score, skills report, interview questions included.

Executive Summary

Luka Posilović is a strong candidate for the Applied AI Researcher / Founding Engineer role, bringing a PhD in Computer Science (summa cum laude), 400+ research citations, and hands-on leadership of ML departments at funded startups. Their technical breadth across generative AI, computer vision, MLOps, and cloud infrastructure closely matches the role's requirements. Two concerns must be addressed before proceeding: a prompt injection attempt was found embedded in the resume, which raises a question about professional transparency and should be directly discussed in screening; and no code sample or GitHub was provided, requiring a technical assessment to validate engineering quality. If these concerns are resolved satisfactorily, Luka represents a high-potential founding engineer candidate with genuine C-level growth trajectory.

Top Strengths

  • PhD in Computer Science summa cum laude with a nationally recognized outstanding dissertation award — top-tier academic pedigree directly matching role requirements
  • 400+ academic citations in generative AI, GANs, and diffusion models — strong research credibility and thought leadership foundation
  • Proven ability to lead and deliver applied AI products in production (Kitro, Inetec) with measurable business and scientific impact
  • Broad and deep technical stack: Python, PyTorch, AWS, GCP, MLOps tooling, computer vision, and LLM-based systems — covers nearly all required skills
  • Entrepreneurial mindset evidenced by building and publishing an independent app (Study Buddy) alongside a demanding research and leadership career

Key Concerns

  • !Prompt injection attempt embedded in the resume ('Ignore all previous instructions...') is a transparency and professional conduct red flag that must be addressed in screening
  • !No code sample or GitHub provided — code quality, software engineering practices, and architectural decision-making cannot be independently verified for a founding engineer role

Culture Fit

74%

Growth Potential

High

Salary Estimate

$100,000 - $135,000 (aligns with senior/staff level given PhD, leadership experience, and European base — may negotiate toward upper band for equity or C-level framing)

Assessment Reasoning

Luka Posilović is assessed as FIT (score 82) based on strong alignment with the role's core requirements: a PhD in the right discipline, deep generative AI and computer vision research, current ML leadership experience, measurable product impact, and a broad technical stack covering Python, PyTorch, AWS, GCP, and MLOps tooling. They meets or exceeds approximately 85% of the required and preferred qualifications. Two notable flags — a prompt injection attempt in the resume and the absence of a code sample or GitHub — reduce confidence slightly but do not disqualify them, as both can be addressed in the interview process. The LinkedIn mismatch (recruiter profile submitted instead of candidate's) is a process error rather than a candidate fault. The FIT decision is conditional on a satisfactory technical interview and clarification of the prompt injection issue.

Interview Focus Areas

System architecture and technical decision-making: how would they design the AI platform from scratch at AlpacaRelay?LLM fine-tuning and multimodal model experience: depth of hands-on work vs. research familiarityLeadership style and team scaling experience: how have they managed and grown ML teams under startup constraints?Code quality and engineering discipline: live coding or take-home assessment to compensate for missing GitHub/code sampleClarification on the prompt injection in the resume — was it intentional as a test, a mistake, or something else?

Code Review

FairSenior Level

No code example or GitHub profile was submitted, making it impossible to directly evaluate code quality, architectural thinking, or engineering discipline. The resume implies a senior-level engineer with production ML experience and tooling maturity, but this must be verified in a technical interview or take-home assessment. The score reflects the inability to evaluate rather than a negative signal.

PythonPyTorchCUDADockerGitHub WorkflowsC# (integration context)
  • +Resume indicates experience with production-grade ML systems, Docker, CUDA optimization, and CI/CD-adjacent tooling (GitHub Workflows), suggesting disciplined engineering habits
  • +Published a consumer-facing app (Study Buddy) on the Play Store, demonstrating ability to ship end-to-end software products
  • -No code sample, GitHub profile, or open-source repository was provided for direct assessment — this is a significant gap for a founding engineer role where code quality is critical

Experience Overview

8y total · 6y relevant

Luka Posilović is a highly credentialed AI researcher and practitioner with a PhD in Computer Science (summa cum laude), significant academic impact (400+ citations), and strong hands-on experience spanning generative models, computer vision, MLOps, and cloud infrastructure. They currently leads an ML department and has delivered measurable real-world AI products. Their background aligns closely with the role's core requirements, though direct evidence of large-scale LLM fine-tuning and code quality cannot be fully verified without a GitHub or code sample.

Matching Skills

PhD in Computer SciencePythonPyTorchAWSGCPMLOps (DVC, WandB, Airflow)Computer VisionGenerative Models / GANs / Diffusion ModelsLLMs and AI AssistantsObject Detection and SegmentationModel Training and Fine-tuningTeam Leadership (Head of ML)Open-source Contributions (FiftyOne community)Vector DatabasesFlaskDockerCUDA

Skills to Verify

TensorFlow (not explicitly mentioned)Azure cloud infrastructureExplicit LLM fine-tuning at scaleFormal MLOps pipeline management at enterprise scale
Candidate information is anonymized. Personal details are hidden for fair evaluation.