Tristan J.

Key Skills

Software

Google Cloud Vertex AI

Other

AWS SageMaker

Data Annotation Tech

CloudFactory

CrowdFlower

Top Subject Matter

Healthcare knowledge assistants and customer support RAG systems

Document processing and workflow automation with human oversight

Multi-agent RAG validation with knowledge graphs

Top Data Types

Image

Text

Computer Code Programming

Top Task Types

Prompt + Response Writing (SFT)

Function Calling

Bounding Box

Segmentation

Classification

Entity (NER) Classification

Polygon

Object Detection

Text Generation

Question Answering

Text Summarization

Fine-tuning

Transcription

Computer Programming/Coding

Data Collection

Freelancer Overview

Senior Gen AI/ML & Prompt Engineer at Saxon AI (05/2023–03/2026). Brings 9+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Google Cloud Vertex AI and Other. Education includes Bachelor of Science, University of Texas, Dallas (2015). AI-training focus includes data types such as Document and labeling workflows including Prompt + Response Writing (SFT), Function Calling, and Evaluation.

Labeling Experience

Gen AI Developer / Team Collaborator — Copilots & Power Platform for Human-Agent Collaborative Automation (Feb 2025)

OtherDocumentFunction Calling

Designed human-agent collaborative AI copilots using RPA + LLM integration to automate document processing and decision support tasks. Implemented agent workflows that coordinate tool calls and human oversight through Teams and Power Apps. Applied LLM orchestration components (LangChain and Semantic Kernel) to enable reliable task execution and reduced manual effort. • Document processing and decision-support copilot design. • Human-AI interaction workflow enablement via Teams/Power Apps. • LLM orchestration using LangChain and Semantic Kernel. • Workflow automation reducing manual steps by 30% in client processes.

2025 - Present

Senior Gen AI/ML & Prompt Engineer at Saxon AI (05/2023–03/2026)

Google Cloud Vertex AIDocumentPrompt + Response Writing (SFT)

Developed and iterated LLM prompt workflows and evaluation loops to improve system interaction efficiency and user satisfaction. Integrated LangChain with CrewAI to support autonomous document retrieval, summarization, and validation for RAG accuracy improvements. Deployed fine-tuned LLMs via Vertex AI endpoints and automated ML workflows for faster iteration and consistent deployment. • Prompt engineering experiments using Llama2/Mistral for healthcare and support assistant use cases. • Multi-agent retrieval/summarization/validation to reduce hallucinations and improve correctness. • Fine-tuned model deployment for chatbot response accuracy gains. • Vertex AI Pipelines automation for reduced model deployment time.

2023 - 2026

LLM Developer (Multi-Agent Architect) — Digital Clerx at SaxonAI (2023)

OtherDocument

Built a vertical multi-agent system for autonomous task execution with human oversight using an LLM-agnostic framework. Implemented a modular RAG pipeline with plug-and-play retrievers (Pinecone/FAISS) and generators (GPT-3/Claude) to support knowledge-grounded outputs. Created a validation agent to fact-check retrieved documents against knowledge graphs, improving cross-agent accuracy. • Orchestrator, Research, and Validation specialized agents. • Retriever/generator swapping without code changes for model flexibility. • Knowledge-graph-based validation of retrieved content. • Achieved 95% task accuracy via cross-agent validation.

2023 - 2023

Education

Work History