danielgaskins.github.io/resume.html at main · danielgaskins/danielgaskins.github.io · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
<!doctype html>
<html lang="en">
  <head>
    <meta charset="UTF-8" />
    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
    <meta
      name="description"
      content="Résumé for Daniel Gaskins, an applied machine learning engineer focused on model evaluation, experimentation, and reliable production systems."
    />
    <title>Daniel Gaskins — Applied Machine Learning Engineer — Résumé</title>
    <link rel="canonical" href="https://danielgaskins.com/resume.html" />
    <link rel="icon" href="./favicon.svg" type="image/svg+xml" />
    <link rel="stylesheet" href="./resume.css" />
  </head>
  <body>
    <nav class="resume-actions" aria-label="Résumé actions">
      <a href="./">← Portfolio</a>
      <a href="./assets/Daniel_Gaskins_Resume.pdf" download>Download PDF</a>
      <button type="button" onclick="window.print()">Print</button>
    </nav>

    <main class="resume">
      <header class="resume-header">
        <h1>Daniel Gaskins</h1>
        <p class="resume-title">Applied Machine Learning Engineer</p>
        <p class="resume-contact">
          <a href="mailto:hello@danielgaskins.com">hello@danielgaskins.com</a>
          <a href="https://danielgaskins.com">danielgaskins.com</a>
          <a href="https://github.com/danielgaskins">github.com/danielgaskins</a>
          <a href="https://www.linkedin.com/in/daniel-gaskins-ml/">linkedin.com/in/daniel-gaskins-ml</a>
          <span>United States · Open to relocation</span>
        </p>
      </header>

      <section aria-labelledby="summary-heading">
        <h2 id="summary-heading">Summary</h2>
        <p>
          Applied AI engineer and founder who has built document-processing, computer-vision,
          agent-evaluation, and business-automation systems. Strongest at turning an unclear problem
          into a testable workflow, finding where it fails, and carrying the result into production.
        </p>
      </section>

      <section aria-labelledby="skills-heading">
        <h2 id="skills-heading">Technical Skills</h2>
        <p><strong>Machine Learning:</strong> Agent evaluation, golden datasets, LLM document extraction, model training and evaluation, experiment design, error analysis, data leakage, train/serve parity, NLP, computer vision, gradient-boosted trees, CLIP</p>
        <p><strong>Frameworks:</strong> PyTorch, TensorFlow, scikit-learn, LightGBM</p>
        <p><strong>Software &amp; Systems:</strong> Python, C++, TypeScript, JavaScript, Node.js, React/Vite, REST APIs, OAuth 2.0, GCP/Firebase, Cloud Functions, Firestore, Cloud Storage, testing, CI/CD</p>
      </section>

      <section aria-labelledby="project-heading">
        <h2 id="project-heading">Selected Technical Projects</h2>
        <article>
          <div class="role-heading">
            <h3>Mendmark | Mutation Testing for Agent Evals</h3>
            <p><a href="https://github.com/danielgaskins/mendmark">GitHub</a> · <a href="https://pypi.org/project/mendmark-evals/">PyPI</a> · 2026</p>
          </div>
          <ul>
            <li>Built an open-source Python tool that plants controlled failures in passing agent traces, reruns a team’s DeepEval metrics, and fails CI when those evals miss wrong tool arguments, repeated side effects, hidden tool errors, or damaged responses.</li>
            <li>Published a versioned golden set with 24 cases, 13 tool contracts, and 263 pinned mutations across ten domains. A response-only evaluator missed 176 faults while a complete trace-and-outcome evaluator caught all 263.</li>
          </ul>
        </article>
        <article>
          <div class="role-heading">
            <h3>lgbm-to-code — Cross-Runtime ML Inference</h3>
            <p><a href="https://github.com/danielgaskins/lgbm-to-code">GitHub</a> · <a href="https://pypi.org/project/lgbm-to-code/">PyPI</a></p>
          </div>
          <ul>
            <li>Built a code generator that turns trained, one-output LightGBM models into dependency-free Python, C++17, or JavaScript raw-score inference.</li>
            <li>Added executed and compiled parity tests across all three runtimes at 1e-12 tolerances, including missing-value routing and explicit rejection of unsupported model types.</li>
          </ul>
        </article>
      </section>

      <section aria-labelledby="experience-heading">
        <h2 id="experience-heading">Professional Experience</h2>

        <article>
          <div class="role-heading">
            <h3>SyncABill — Founder &amp; Applied AI Engineer</h3>
            <p>2025–Present</p>
          </div>
          <ul>
            <li>Built and deployed a React/Vite invoice application with TypeScript and Node.js services on Google Cloud/Firebase, using Firestore for workflow state and Cloud Storage for source documents.</li>
            <li>Connected structured document extraction to deterministic checks, controller-configured review rules, and QuickBooks Online or Xero.</li>
            <li>Completed Google’s CASA Tier 2 security assessment and brought the product to SOC 2 readiness.</li>
          </ul>
        </article>

        <article>
          <div class="role-heading">
            <h3>Casabauhaus — Founder &amp; Operator</h3>
            <p>2022–Present</p>
          </div>
          <ul>
            <li>Operate a vintage-furniture business spanning sourcing, pricing, merchandising, sales, and delivery, supported by agent and API workflows for reporting, social media, seasonality analysis, and Shopify.</li>
            <li>Developed furniture-valuation and mid-century classification models used in sourcing decisions, including a fine-tuned CLIP model built before modern vision-language models were widely practical.</li>
          </ul>
        </article>

        <article>
          <div class="role-heading">
            <h3>Perfsy — Founder &amp; Machine Learning Engineer</h3>
            <p>2021–2022</p>
          </div>
          <ul>
            <li>Built a computer-vision pipeline to extract structured fields from scanned DMV vehicle titles.</li>
            <li>Normalized scan orientation with OpenCV and experimented with character primitives, autoencoders, convolutional features, and DNN classifiers.</li>
            <li>Created data and evaluation workflows to inspect recognition failures and improve the pipeline.</li>
          </ul>
        </article>

      </section>

      <section aria-labelledby="education-heading">
        <h2 id="education-heading">Education</h2>
        <div class="role-heading education">
          <h3>University of California, San Diego</h3>
          <p>Condensed Matter Physics | 2013–2016</p>
        </div>
      </section>

    </main>
  </body>
</html>