Skip to content
Tyler Jackson
Portrait of Tyler Jackson

HELLO

David (Tyler) Jackson

AI/ML engineer working in computer vision, multi-modality, and edge deployment

AI/ML engineer and M.S. Computer Science candidate (AI/ML specialization) with hands-on experience in multi-modality, computer vision, object detection, contrastive representation learning, and deployment of models to edge GPU hardware (Jetson Orin AGX, Jetson Nano, DGX Spark).

M.S. Computer Science (AI/ML) @ UCF · AI Solutions Architect @ Command Post Technologies · Orlando, FL

Download résuméSee projects

EXPERIENCE

Work, research, and study

  1. Aug 2026 – Present

    Command Post Technologies

    AI Solutions Architect

    Orlando, FL

    • Designing a multimodal deep learning system that refines low-quality geospatial vector features against aerial and satellite imagery, using per-modality encoders (imagery, vector geometry, geocoordinates, depth) contrastively aligned with CLIP-style methods after GeoCLIP/UniGeoCLIP.
    • Architecting the company's enterprise generative AI rollout on AWS Bedrock: model access and provisioning, Bedrock Guardrails for content and policy filtering, and the shared service layer other teams build against.
    • Developing a CUI-compliant knowledge base and MCP server that exposes controlled document retrieval to Bedrock-hosted models, with an in-server ingestion pipeline handling document types such as PDF, DOCX, and PPTX.
    • Packaged and automated the team's AWS deployment tooling as an internal Python distribution; diagnosed and remediated the distribution and access failures blocking multi-host rollout.
    • Python
    • multi-modal contrastive learning (CLIP-style encoders)
    • Geospatial ML
    • AWS
    • AWS Bedrock
    • MCP
    • RAG
    • Document ingestion (PDF, DOCX, PPTX)
    • Python packaging

  2. Jun 2026 – Aug 2026

    Command Post Technologies

    AI/ML Engineer Intern

    Orlando, FL

    • Project Lead and Full Stack Engineer on an Agentic Marketing AI Platform, owning Frontend, Backend, Database, MCP Deployment, and AI Deployment; cut manual creation of Social Media Posts, Job Postings, and Newsletter down by 70%.
    • Interviewed the marketing team to scope pain points and source grounding content for a LangGraph RAG pipeline, and built in-UI rating and feedback capture that persists scored response pairs as a labeled preference dataset.
    • Built the Backend on a company-hosted NVIDIA DGX Spark utilizing Python 3.12 with Fast API hosting vLLM running GPT-OSS-20b for RAG retrieval and main document drafting, as well as a Stateless MCP Server hosted on a P100 hosting Ollama running a LLaVA-3 b model for image analysis.
    • Python
    • FastAPI
    • vLLM
    • GPT-OSS-20b
    • LangGraph
    • RAG
    • MCP
    • Ollama
    • LLaVA
    • DGX Spark
    • Full-stack
    • Preference datasets
    Two people standing behind a teal Command Post Technologies table at a conference booth, with CPT and CADRE banners behind them.Eight people holding framed certificates in front of a Command Post Technologies logo on an office lobby wall.

  3. Jan 2025 – Jun 2026

    Growing Martin Minds

    Google Cloud Developer (Part-Time)

    • Architected an automated backend for payroll, invoicing, and tutor log tracking; built Apps Script automations for document generation and processing.
    • Designed data pipelines and validation logic that eliminated manual entry error paths.
    • Google Cloud Services
    • Google Apps Script
    • Google Workspace
    • Data pipelines

  4. Aug 2020 – Jan 2025

    Element Construction Services of Orlando (ECSO)

    General Office Specialist / Estimator & Operations Assistant (Part-Time)

    Orlando, FL

    • Configured the Buildertrend platform for company-wide adoption, including template job folder structures, workflow calendars, and bid management systems.
    • Designed and maintained Google Sheets used for estimating, budgeting, and financial tracking — several of which remain in daily use five years later.
    • Developed a robust multi-sheet bill approval system using Google Apps Script, automating approval flows between departments.
    • Integrated Google Sheets with QuickBooks via REST API to automate bill transfers and reduce manual accounting steps.
    • Introduced automation that reduced manual accounting effort and improved bill approval turnaround time by over 50%.
    • Implemented and maintained project boards in Monday.com, creating automated workflows using built-in scripting tools to streamline communication and progress tracking.
    • Drafted minor house plans, renovation layouts, and detailed schematics using AutoCAD-Lite.
    • Buildertrend
    • Google Sheets
    • Google Apps Script
    • QuickBooks REST API
    • REST APIs
    • Monday.com
    • AutoCAD-Lite

SELECTED WORK

What I’ve built

All projects

SKILLS

What I work with

  • Machine Learning & Computer Vision

    Object detection, multi-modal contrastive learning, retrieval, and transformer inference optimization.

    • Object detection (YOLO)
    • multi-modal contrastive learning (CLIP-style encoders)
    • model evaluation and benchmarking
    • RAG
    • Light-RAG
    • LoRA / QLoRA fine-tuning
    • LangChain
    • LangGraph
    • transformer inference optimization (vLLM, SGLang, PagedAttention, FlashAttention)

  • Edge & Accelerated Compute

    NVIDIA Jetson and DGX Spark hardware, GPU inference profiling, and offline deployment.

    • NVIDIA Jetson Orin AGX
    • Jetson Nano
    • DGX Spark
    • GPU inference profiling
    • offline/air-gapped deployment

  • Full-Stack & Cloud

    Languages, frameworks and libraries, and cloud and infrastructure tooling.

    • Python
    • C/C++
    • Java
    • TypeScript/JavaScript
    • SQL
    • PyTorch
    • TensorFlow
    • Keras
    • React
    • React Native
    • Flutter
    • Node.js
    • Express.js
    • Flask
    • Google Cloud Services
    • Azure
    • FastAPI
    • REST APIs
    • Linux
    • Git
    • Docker
    • Kubernetes
    • Neo4j
    • PostgreSQL
    • virtual machines

  • Mathematics

    Linear algebra, matrix methods, probability and statistics, and transform methods.

    • Linear algebra
    • matrix methods
    • probability and statistics
    • convolution and transform methods
    • trigonometry

  • Tools & Methodologies

    Planning, tracking, and development tooling.

    • Jira
    • Confluence
    • Agile
    • Linear
    • Claude Code
    • Cursor

LEADERSHIP

Teaching and volunteering

  • F.E.A.S.T — Foundation Exam review sessions

    Foundation Exam review sessions (F.E.A.S.T) for UCF CS students covering C, data structures, and algorithms.

    Three presenters at the front of a classroom in front of a projected slide titled 'What was our strategy to be successful on the FE?', speaking to a seated audience of students.
    A Foundation Exam review session.

  • 3rd-grade AI presentation

    Annually deliver an introductory AI presentation to a 3rd-grade class on modern-day algorithms.

  • LightOrlandoHungerProject

    Assembly line to create a million meals for those in need.

CONTACT

Say hello

The fastest way to reach me is email.

davidtjackson1029@gmail.com
Email meLinkedInGitHubDownload résumé