Robert P.

Key Skills

Software

No software listed

Top Subject Matter

Local LLM architecture

Fine-tuning Domain Expertise

and evaluation for cybersecurity-focused agentic applications

Top Data Types

Text

Document

Top Task Types

Fine-tuning

Freelancer Overview

Founder & Lead Engineer — trained, fine-tuned, evaluated, and deployed a local language model (Comet) for an agentic IDE. Brings 9+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Ollama. Education includes Bachelor of Science, Colorado State University Global (CSU Global) (2024). AI-training focus includes data types such as Text and labeling workflows including Fine-tuning and Fine Tuning.

Labeling Experience

Founder & Lead Cybersecurity/Software Engineer - Velkron Technologies

Text

Led the architecture and development of an agentic IDE for local file system and terminal automation, reducing reliance on external cloud services. Fine-tuned and evaluated a large local language model while mitigating algorithmic vulnerability vectors through structured prompt engineering and log audits. Designed and deployed secure architectures for decentralized commercial platforms, implementing row-level security and IAM integration for third-party APIs. • Architected local agentic IDE workflows and isolated proprietary computation • Trained and evaluated a 100-million parameter local language model with security-focused audits • Implemented Supabase row-level security and strict IAM integration protocols • Built continuous integration scanning pipelines for web vulnerability diagnostics

2024 - Present

Founder & Lead Engineer — trained, fine-tuned, evaluated, and deployed a local language model (Comet) for an agentic IDE (Velkron Code).

TextFine-tuning

You trained and fine-tuned a local language model and evaluated it using systematic audits of evaluation logs. You used prompt engineering to reduce algorithmic vulnerability vectors and improve safe, reliable behavior for agentic workflows. Work focused on preparing and validating model outputs rather than manual annotation. • Fine-tuned a 100M parameter local language model • Conducted structured evaluation log audits • Applied rigorous structural prompt engineering • Deployed the model for local agentic IDE use

2024 - Present

Education

Work History