Arjun Kulkarni
Product / agent engineer building AI systems you can trust in production.
Works best in a product role at a fast-moving startup.
Interests · trustworthy and secure AI, physical AI, nanotechnology
New York City · Open to NYC / SF / Remote
Education
Georgia Institute of TechnologyJun 2026 – Jan 2027
M.S. Computer Science & Artificial Intelligence
University of Illinois at Urbana-ChampaignSep 2022 – May 2026
B.S. Computer Science & Materials Engineering
Experience
BrekenAug 2026 – Present
Founding Engineer · Production simulation
Production simulation that finds failures before customers do
Built a simulator that recreates a company's software, finds the bugs customers would hit, and drafts the fix.
Backed by
IBMJune 2026 – Aug 2026
Applied AI Engineer · Agent reliability research
Composable agent skills under IBM Fellow Dr. Kush R. Varshney
Built reusable AI skills with the Mellea Skills Compiler and prototyped reliable agent workflows.
NestléJune 2026 – Aug 2026
Applied AI Engineer · AI for food R&D
LLM retrieval over a 2M+ recipe co-occurrence corpus
Built retrieval and data pipelines pairing formulation signals with semantic search.
FitFoFeb 2026 – June 2026
CTO / AI Systems · Creator fitness coaching
40K users · $350K GMV · $500K raised
Led the platform as CTO, from architecture to an AI coach grounded in creator content.
HandshakeFeb 2026 – May 2026
Software Engineer, AI Evaluation · Multimodal AI evaluation
Training-data QA for state-of-the-art vision models
Built Kubricks for video ingestion and transcription, and evaluated models against gold annotations.
SideShiftMay 2026
Software Engineer · Messaging and contracts
~95% fewer database reads and writes on messaging
Added Redis + TanStack Query caching and built in-app contract signing.
UCSF HealthApr 2026 – May 2026
Software Engineer Intern · Clinical AI
Agentic EHR supporting 20+ patients daily
Built the Flask pipelines behind Congruence's SOAP notes and patient context.
Curator.toJan 2024 – Dec 2025
Founding Engineer · Agentic e-commerce
Architected the multi-tenant TypeScript backend and its tool-based agents.
Backed by
UC San Diego HealthJan 2025 – Aug 2025
Software Engineer Intern · Clinical RAG
Dietary recommendations +22% more accurate for 200+ patients
Built LangGraph + QDrant clinical reasoning over patient history and allergies.
AbbVieDec 2024 – Jun 2025
Software Engineer Intern · Oncology search
ETL over 120K+ reports · R&D lookup ~3x faster
Built FastAPI + FAISS semantic search for clinical trial documents.
RumorOct 2025 – Jan 2026
Software Engineer Intern · Real-time events
~85% faster search for 50K+ users
Cut event-search latency with Redis caching and shipped React Native features.
Research
Undergraduate Teaching Assistant
Jan 2023 – Apr 2023
Siebel School of Computing and Data Science, University of Illinois
- Project
- TA for CS124 (Java Programming) and CS128 (Introduction to C++). Held office hours, gave one-on-one help, and graded assignments with detailed feedback.
- Outcome
- Helped students through coursework and projects in two intro CS courses.
JavaC++Data StructuresAlgorithmsCode ReviewSustainable Concrete with ML
Jan 2024 – May 2024
Civil and Environmental Engineering, University of Illinois
- Project
- Analyzed lab data and designed a new concrete mixture with less clinker in the cement, using Random Forest models to predict how composition affects durability.
- Outcome
- 10% lower carbon emissions.
PythonRandom ForestScikit-learnPandasNumPyAutonomous EV Routing
Jan 2024 – Jan 2025
Illinois AI Group, Electrical & Computer Engineering, University of Illinois
- Project
- Put a reinforcement learning agent inside a model predictive control (MPC) system that routes electric vehicles, tested in SUMO traffic simulations.
- Outcome
- 18% less traffic delay and 25% better control accuracy.
RLMPCSUMOROSC++PythonAI Dietitian
Jan 2024 – Jan 2025
Coordinated Science Laboratory (CSL), University of Illinois
- Project
- Led DevOps and ML for an AI dietitian with Food Network Chef James Briscione and White House Fellow Lav Varshney: a LangGraph + RAG pipeline over a QDrant vector database. Later merged into the UC San Diego Health project.
- Outcome
- 22% more accurate dietary recommendations.
LangChainLangGraphRAGQDrantOpenAIFlaskPythonHuman–AI Interaction Platform
Feb 2025 – Apr 2025
Beckman Institute for Advanced Science and Technology, University of Illinois
- Project
- Built a real-time React + Flask platform where multiple people talk to an LLM at once, with WebSockets and analytics logging for research experiments.
- Outcome
- Ran studies with 40+ subjects and zero lost messages.
ReactFlaskWebSocketsPostgreSQLLLMNanofluidic Energy Harvesting
Sep 2025 – Dec 2025
Illinois Quantum Information Science and Technology Center (IQUIST), University of Illinois
- Project
- Modeled ion-electron Coulomb drag to predict voltage and current in silicon nanochannels, then designed carbon nanotube channels to harvest more energy.
- Outcome
- Channel designs aimed at higher energy-harvesting efficiency, building on work published in Nano Energy (2023).
PythonCOMSOLMATLABNanofluidicsDevice Physics
Selected work
Skills
- Languages (Core)
- PythonTypeScriptJavaScriptC++SQLJava
- Frameworks
- ReactReact NativeNext.jsFlaskFastAPINode.jsExpressDjangoSpring Boot
- AI/ML
- PyTorchTensorFlowLangChainRAGRLLLM OpsOpenAI AssistantsGeminiVector SearchLangGraph
- Cloud/DevOps
- AWSECSFargateDigitalOceanDockerTerraformKubernetesGitHub ActionsLinuxCI/CDJenkinsGitLab CI/CD
- Tools
- PostgresMongoDBRedisKafkaTemporalClickHousePineconeTurbopufferROSSUMOCOMSOLGitJiraLinearAsana
Clubs & community
IEEE · Talks on hardware, quantum, and ML.
Illinois Business Consulting · Go-to-market with CIOs and CTOs.
SIGPwny · Secure UART and Rust firmware for a medical device.
NOBE Tech Lab · Shipped sites and tools for 30+ clients.