Jakob Nordhagen
Software Engineer
jnordhagen1@gmail.com · linkedin.com/in/jakobnordhagen · github.com/jnordhagen
Experience
Software Engineer — Celerity Solutions
Aug. 2024 – Present · Palo Alto, CA- Reduced AI feature latency 55% by refining chained prompt architecture and resolving critical issues affecting 1700+ production users
- Accelerated month-segmented sales data ingestion 10x by rewriting core SQL logic in the materialization process
- Delivered customer movement and pricing insights to cross-functional teams, launching custom-built app modules
Software Engineer Intern — Celerity Solutions
May 2024 – Aug. 2024 · Palo Alto, CA- Deployed new AI feature for building data reports with 92% on-task accuracy, reducing manual work for users
- Cut hallucination rate 40% and increased AI feature scope by designing new chained-prompt architecture with parallel task processing
- Unlocked granular LLM knowledge of 250,000+ products, categories, stores, and brands by implementing and integrating OpenSearch-based RAG pipeline and vector database
Graduate Research Assistant — Stanford Trustworthy AI Research Lab
Jan. 2024 – Apr. 2024 · Stanford, CA- Implemented custom LLM training loop with on-the-fly backtranslation for math autoformalization, enabling cost-efficient fine-tuning of 175M–2B parameter models
- Co-authored research paper submitted to ICLR 2025 on fine-tuning techniques for LLM autoformalization
Education
M.S. Computer Science (AI/ML) — Stanford University
Sept. 2022 – June 2024B.S. Computer Science (Systems) — Stanford University
Sept. 2019 – June 2024Skills
Languages: Python, C++, C, TypeScript, SQL, CUDA, Go, Julia
ML/AI Tools: PyTorch, TensorFlow, Hugging Face, LangChain, Pandas, NumPy
Cloud/DevOps: AWS (Bedrock, Lambda, Redshift), Docker, GCP, CloudWatch
Frameworks/Databases: React, Node.js, Next.js, MongoDB, OpenSearch, Pinecone
Projects
See /projects for the full list.