{"id":43231,"date":"2026-02-04T09:33:11","date_gmt":"2026-02-04T14:33:11","guid":{"rendered":"https:\/\/www.bu.edu\/cise\/?page_id=43231"},"modified":"2026-02-07T11:51:54","modified_gmt":"2026-02-07T16:51:54","slug":"cise-seminar-vassilis-digalakis-jr-boston-university","status":"publish","type":"page","link":"https:\/\/www.bu.edu\/cise\/cise-seminar-vassilis-digalakis-jr-boston-university\/","title":{"rendered":"CISE Seminar: Vassilis Digalakis Jr, Boston University"},"content":{"rendered":"

Date: February 13, 2026
\nTime: 3:00pm \u2013 4:00pm
\nLocation: 665 Commonwealth Ave., CDS 1101<\/p>\n

\"\"Boston University
\nAssistant Professor of Operations & Technology Management\u00a0<\/span>
\nBoston University<\/h4>\n

ML Compass: Navigating Capability, Cost, and Compliance Trade-offs in AI Model Development<\/strong><\/p>\n

In this talk, I will discuss how organizations should select among competing AI models when user utility, deployment costs, and compliance requirements jointly matter. The talk relies on joint work with Ramayya Krishnan (CMU), Gonzalo Martin Fernandez (UPC), Agni Orfanoudaki (Oxford). The paper is available at https:\/\/www.arxiv.org\/abs\/2512.23487.<\/a><\/div>\n
<\/div>\n
Widely used capability leaderboards do not translate directly into deployment decisions, creating a capability — deployment gap; to bridge it, we take a systems-level view in which model choice is tied to application outcomes, operating constraints, and a capability-cost frontier. We develop ML Compass, a framework that treats model selection as constrained optimization over this frontier. On the theory side, we characterize optimal model configurations under a parametric frontier and show a three-regime structure in optimal internal measures: some dimensions are pinned at compliance minima, some saturate at maximum levels, and the remainder take interior values governed by frontier curvature. We derive comparative statics that quantify how budget changes, regulatory tightening, and technological progress propagate across capability dimensions and costs. On the implementation side, we propose a pipeline that (i) extracts low-dimensional internal measures from heterogeneous model descriptors, (ii) estimates an empirical frontier from capability and cost data, (iii) learns a user- or task-specific utility function from interaction outcome data, and (iv) uses these components to target capability-cost profiles and recommend models. We validate ML Compass with two case studies: a general-purpose conversational setting using the PRISM Alignment dataset and a healthcare setting using a custom dataset we build using HealthBench. In both environments, our framework produces recommendations — and deployment-aware leaderboards based on predicted deployment value under constraints — that can differ materially from capability-only rankings, and clarifies how trade-offs between capability, cost, and safety shape optimal model choice.<\/div>\n

 <\/p>\n

Vassilis Digalakis Jr<\/strong> is an Assistant Professor of Operations & Technology Management at Boston University\u2019s Questrom School of Business.\u00a0He completed his Ph.D. in Operations Research at MIT, advised by Prof. Dimitris Bertsimas, and holds a Diploma in Electrical and Computer Engineering from the Technical University of Crete, Greece.\u00a0He has also spent time at HEC Paris, France as an Assistant Professor and the Alexa Natural Language Understanding team at Amazon Science as a Research Scientist Intern.<\/p>\n

His research is on trustworthy AI. He studies how to develop AI\/ML models that are trustworthy (e.g., interpretable, robust, stable); how to leverage these properties to increase adoption; and how to deploy models in real operational settings, especially in healthcare and sustainability. His work is both methodological\u2014at the interface of machine learning and optimization\u2014and applied, including collaborations with FEMA on vaccine allocation and OCP in renewable energy capacity planning. His\u00a0research has been published in journals such as\u00a0Operations Research\u00a0and\u00a0Manufacturing & Service Operations Management, and has received awards, including the\u00a0INFORMS Pierskalla Award and the Harold Kuhn Award.<\/p>\n

Faculty Host: <\/strong>Ayse Coskun
\nStudent Host:<\/strong> Jiatong Guo<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"

Date: February 13, 2026 Time: 3:00pm \u2013 4:00pm Location: 665 Commonwealth Ave., CDS 1101 Boston University Assistant Professor of Operations & Technology Management\u00a0 Boston University ML Compass: Navigating Capability, Cost, and Compliance Trade-offs in AI Model Development In this talk, I will discuss how organizations should select among competing AI models when user utility, deployment […]<\/p>\n","protected":false},"author":18553,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"page-templates\/no-sidebars.php","meta":[],"_links":{"self":[{"href":"https:\/\/www.bu.edu\/cise\/wp-json\/wp\/v2\/pages\/43231"}],"collection":[{"href":"https:\/\/www.bu.edu\/cise\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.bu.edu\/cise\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.bu.edu\/cise\/wp-json\/wp\/v2\/users\/18553"}],"replies":[{"embeddable":true,"href":"https:\/\/www.bu.edu\/cise\/wp-json\/wp\/v2\/comments?post=43231"}],"version-history":[{"count":8,"href":"https:\/\/www.bu.edu\/cise\/wp-json\/wp\/v2\/pages\/43231\/revisions"}],"predecessor-version":[{"id":43244,"href":"https:\/\/www.bu.edu\/cise\/wp-json\/wp\/v2\/pages\/43231\/revisions\/43244"}],"wp:attachment":[{"href":"https:\/\/www.bu.edu\/cise\/wp-json\/wp\/v2\/media?parent=43231"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}