Jieren Deng, Ph.D. (邓杰仁, 鄧傑仁)
what doesn't kill you makes you stronger
I am a Lead AI Engineer with research focus at Capital One AI Foundation team, working on next-generation AI systems. I obtained my Ph.D. from the School of Computing at the University of Connecticut (UCONN), advised by Professor Derek Aguiar in the EL GATO Lab.
Outside work, I share essays on AI and engineering on LinkedIn and build open source projects. I like turning practical ideas into tools that make everyday workflows simpler, faster, and more useful.
Industry Impact
Microsoft (2024-2026)
- Multimodal AI & Vision-Language Models: Leading multiagent and multimodal interactive maps experience for IMAIA (accepted CVPR 2026), focusing on spatial understanding, geospatial intelligence, and cross-modal reasoning; 3 patents filed.
- Agentic AI & Multi-Agent Systems: Leading AIOps evolution for Azure Core with agent-based observability systems for autonomous agents, incident triaging, and intelligent reasoning; one research paper accepted FSE 2026.
Baidu Research USA (2022-2023)
- Knowledge distillation and diffusion model research (ICASSP 2024 Oral, ICIP 2023).
Selected Publications
- CVPR 2026 - IMAIA: Interactive Maps AI Assistant for Travel Planning and Geo-Spatial Intelligence (Microsoft AI)
- FSE 2026 - An Agentic Framework for Triaging Incidents in Production Cloud Infrastructure (Microsoft Azure AIOps)
- ICASSP 2024 (Oral) - GBSD: Generative Bokeh with Stage Diffusion (Baidu Research)
- EMNLP 2021 (Oral) - A Secure and Efficient Federated Learning Framework for NLP
- EMNLP 2021 - TAG: Gradient Attack on Transformer-based Language Models
Notes on AI & Engineering
Research notes on AI, engineering, and quantitative experiments. Also find my essays on LinkedIn.
November 2026 · Research notes
Congress Signal Lab: Reading Portfolios Like Documents
A TF-IDF-inspired experiment returned +32.26% versus SPY's +11.05% in its original annual backtest. The idea, return drivers, and unresolved audit findings behind that historical result.
August 2026 · LinkedIn
Agentic Systems in Practice: Latency, Reasoning, Skills, and Token Economics
A practical perspective on how workflow-enhancement, open-ended consumer, and enterprise agents require different trade-offs across latency, reasoning, skills, and cost.
GitHub Open Source
A few public projects that reflect how I like to build: practical automation, product-minded tools, and systems that stay useful after the first demo.
Public Repository
CongressSignalLab
An offline Python research toolkit that treats congressional buying histories as documents and stock tickers as tokens, using TF-IDF-inspired ranking to study unusual investment behavior.
Original annual backtest: +32.26% vs. SPY +11.05%. Historical, in-sample results with known defects; corrected results pending.
Public Repository
uscis-case-API-monitor
A USCIS case monitor that combines Playwright-based authentication, API polling, JSON diffing, and notifications to reduce manual status checking.
Personal pain point turned into a durable automation system.
Public Repository
CreditCardMaster
An open source credit card recommendation tool for comparing options and surfacing the best fit through a simpler decision flow.
A practical consumer-facing project with product-style UX thinking.
News
| Oct 2026 | Serving as Agentic AI Area Chair for the 3rd Synthetic & Adversarial ForEnsics (SAFE) Workshop at AAAI 2027, helping lead the review and program decisions for agentic AI submissions. |
|---|---|
| Apr 2026 | Joined Capital One AI Foundation team as a Lead AI Engineer with research focus on next-generation AI systems and intelligence! |
| Mar 2026 | FSE 2026 accepted! An Agentic Framework for Triaging Incidents in Production Cloud Infrastructure - Exploring how AI agents can automate incident triage in large-scale cloud systems. |
| Feb 2026 | CVPR 2026 accepted! IMAIA: Interactive Maps AI Assistant for Travel Planning and Geo-Spatial Intelligence - A new paradigm for conversational mapping with natural language interaction. Read the paper → |
| Jul 2025 | New publication! Distilling Adversarial Robustness Using Heterogeneous Teachers in Procedia Computer Science. Leveraging diverse teacher models to improve robustness through knowledge distillation. Read more. |
| Jun 2025 | Certifying Adapters: Enabling and Enhancing the Certification of Classifier Adversarial Robustness accepted by IJCNN 2025! Check it out here. |
| Dec 2024 | Successfully defended my Ph.D. thesis at UConn! Grateful for the support and excited for what’s next. |
| Jun 2024 | Joined Microsoft AI as a Data & Applied Scientist! Excited to work on cutting-edge AI research and applications. |
| May 2024 | Awarded Predoctoral Fellowship (2024) from UConn CSE - Honored for research contributions and academic achievements! |
| Dec 2023 | GBSD: Generative Bokeh with Stage Diffusion accepted as an Oral Presentation at ICASSP 2024! A novel approach to generating artistic bokeh effects. Read on arXiv |
| Nov 2023 | Appointed as Guest Lecturer for UConn Graduate Course CSE 5825 - Bayesian Machine Learning! |
| Sep 2023 | Invited to serve as a reviewer for SDM 2024 - SIAM Conference on Data Mining. |
| Sep 2023 | Serving as a reviewer for ICASSP 2024 - The IEEE International Conference on Acoustics, Speech and Signal Processing. |
| Sep 2023 | Invited to serve as a reviewer for AAAI 2024 - Helping advance the field through rigorous peer review. |
| Jul 2023 | Serving as a reviewer for ICML AdvML-Frontiers 2023 and EMNLP 2023 - Contributing to the community through peer review. |
| Jun 2023 | Joined Microsoft Maps AI team for a Data & Applied Scientist internship - Working on next-generation AI for spatial understanding. |
| Jun 2023 | SSSD: Single-Stage Semantic Diffusion accepted by ICIP 2023 - A new approach to image processing with diffusion models. Read on arXiv |
| May 2023 | Awarded Predoctoral Fellowship (2023) from UConn CSE - Grateful for continued support in research! |
| Jan 2023 | Honored to accept the invitation as a Program Committee (PC) member for KDD 2023! |
| May 2022 | Honored to accept the Predoctoral Fellowship from UConn CSE - Grateful for the support in research endeavors! |
| May 2022 | Started research internship at Baidu USA - Exploring advanced AI research in diffusion models and knowledge distillation. |
| Apr 2022 | SAPAG accepted by IJCNN-WCCI 2022! Novel framework for learning with partial membership. Read on IEEE Xplore |
| Nov 2021 | Paper on weight pruning optimization accepted by DATE 2022 - Advances in design automation and testing. Read on IEEE Xplore |
| Aug 2021 | FL-DISCO: Federated Learning with Distributed Collaborative Optimization accepted by ICCAD 2021 - Advancing distributed machine learning. Read on IEEE Xplore |
| Aug 2021 | A Secure and Efficient Federated Learning Framework for NLP accepted by EMNLP 2021 as an Oral Presentation! Privacy-preserving collaborative NLP training. Read on arXiv |
| Aug 2021 | TAG: Targeted Adversarial Gradients on Transformer-based Language Models accepted by EMNLP 2021! Exploring vulnerabilities in pre-trained language models. Read on arXiv |
| Jun 2021 | Internship at Oak Ridge National Laboratory (ORNL) - Collaborating on cutting-edge scientific research with national labs. |
| Feb 2021 | TinyADC: Model compression for analog-to-digital converters accepted by DATE 2021 with a Best Paper Nomination! Read on IEEE Xplore |
| Nov 2020 | ESMFL: Efficient Secure Multi-party Federated Learning accepted by NeurIPS Workshop 2020 (SpicyFL) - Privacy-preserving collaborative machine learning. Read on arXiv |
| Aug 2020 | Started my Ph.D. journey at University of Connecticut (UConn) in the Department of Computer Science and Engineering! |
| Aug 2020 | Awarded Cigna Graduate Fellowship from UConn CSE - Recognition for academic excellence and research potential. |
| Aug 2020 | Paper accepted by SOCC 2020 - The 33rd IEEE International System-on-chip Conference! Exploring system designs and optimizations. Read on IEEE Xplore |
| Jun 2020 | iPRIVATES published in Briefings in Bioinformatics 2020 - Privacy-preserving analysis of genomic data. Read on Oxford Academic |