DoorDash
May 2026 - Aug 2026Machine Learning Engineer Intern — Agentic AI Research
Applied research on a self-healing production agent to improve Dasher productivity, connecting failure analysis with automated prompt optimization.
Shreeyash Gowaikar
Hi there! I'm a Computer Science master's student at Georgia Tech, specializing in Machine Learning. I work with Prof. Kai Wang in the AI for Social Impact Team.
My research has taken me from responsible AI at UdeM to agentic systems and code intelligence at Microsoft Research, and applied research on self-healing production agents at DoorDash. Across these settings, a recurring challenge has been making AI dependable in practice.
Those experiences shaped my interest in AI robustness and safety. Today, I focus on fine-tuning robustness, bringing diffusion models and LLMs together to understand how adaptation affects model behavior and how to preserve safety and reliability.
Research & Work History
Applied Research in Agentic AI
AI Robustness & Reliability
AI for Software Eng. and AI for Manufacturing
AI for Urban Planning
AI for Service Management
AI Evaluations (analogical aptitude)
Undergraduate Studies
Latest Updates
📄 Paper 👨🏽💻 Work 🎓 Academic
May 2026
Joined DoorDash!
Joined as a Machine Learning Engineer Intern in Sunnyvale for applied research in Agentic AI: a Dasher copilot and a self-healing pipeline for evaluation, failure analysis, and prompt optimization to improve Dasher productivity.
Sept 2025
NeurIPS'25 Workshop Acceptance
Our work from Microsoft Research, COSMIR, was accepted at the NeurIPS'25 Foundations of Reasoning in Language Models Workshop!
Aug 2025
Started my Grad School!
Joined Georgia Tech for my Masters. Go Jackets!
May 2025
AIES'25 Acceptance
Undergraduate thesis accepted at the AAAI/ACM AI Ethics and Society Conference!
Dec 2024
AAAI'25 Workshop Acceptance
Work from Microsoft Research accepted at the AAAI'25 Workshop on AI to Accelerate Science and Engineering.
Nov 2024
Joined Microsoft Research!
Working on AI for Large-Scale Software Engineering Tasks.
Jul 2024
Research Intern @ MSR India
Generative AI for the Energy domain.
June 2024
Graduated!
BITS Pilani Goa Campus (CGPA 9.15/10) with a minor in Data Science.
May 2024
CVPR'24 Workshop Paper
First author work (AI-EDI-SPACE) accepted at the Workshop on Responsible Data.
May 2026
Joined DoorDash!
Joined as a Machine Learning Engineer Intern in Sunnyvale for applied research in Agentic AI: a Dasher copilot and a self-healing pipeline for evaluation, failure analysis, and prompt optimization to improve Dasher productivity.
Sept 2025
NeurIPS'25 Workshop Acceptance
Our work from Microsoft Research, COSMIR, was accepted at the NeurIPS'25 Foundations of Reasoning in Language Models Workshop!
Aug 2025
Started my Grad School!
Joined Georgia Tech for my Masters. Go Jackets!
May 2025
AIES'25 Acceptance
Undergraduate thesis accepted at the AAAI/ACM AI Ethics and Society Conference!
Show older updates
Dec 2024
AAAI'25 Workshop Acceptance
Work from Microsoft Research accepted at the AAAI'25 Workshop on AI to Accelerate Science and Engineering.
Nov 2024
Joined Microsoft Research!
Working on AI for Large-Scale Software Engineering Tasks.
Jul 2024
Research Intern @ MSR India
Generative AI for the Energy domain.
June 2024
Graduated!
BITS Pilani Goa Campus (CGPA 9.15/10) with a minor in Data Science.
May 2024
CVPR'24 Workshop Paper
First author work (AI-EDI-SPACE) accepted at the Workshop on Responsible Data.
In practice
Experience
Microsoft Research
Nov 2024 - Aug 2025Pre-doctoral Researcher
Multi-agent reasoning and structured memory for code intelligence across large enterprise repositories, including COSMIR.
University of Montreal
Jul 2023 - Dec 2023Undergraduate Student Researcher
Responsible AI research on equity in public spaces and fairness in preference learning, published at CVPR workshops and AIES.
On paper
Selected Research
COSMIR: Chain Orchestrated Structured Memory for Iterative Reasoning
Structured memory and verifiable sub-steps for reasoning over long inputs.
NeurIPS'25 Workshop on Foundations of Reasoning in Language Models (FoRLM@NeurIPS'25) Read paperFrom Efficiency to Equity: Measuring Fairness in Preference Learning
Fairness metrics for preference learning, informed by inequality measures and Rawlsian justice.
AAAI/ACM AI Ethics and Society'25 (AIES'25) Read paper
After hours
Beyond Code
A little beyond the keyboard
When I’m not coding, you can find me stargazing, taking photos, or petting cats.
Our only Moon, captured through a Dobsonian telescope.
Explore my photography →Small world, good company
A Little Corner of the Internet
My counter saves approximate visit totals and grouped locations, not raw IP addresses or individual visit histories. A temporary per-tab flag avoids recounting. Cloudflare processes the network requests.
Privacy details → Drag to rotate · arrow keys work too
Thanks for stopping by 👋
It’s a small world. I’m glad our corners of it crossed. The cat thinks it’s a toy.
Loading visitor statistics…
- Visits
- Countries / regions
Location dots appear as visits arrive.