# Ross Klein

ross@rossklein.com | (516) 729-2584 | [rossklein.com](https://rossklein.com/) | [github.com/rossklein](https://github.com/RossKlein/) | [linkedin.com/in/ross-klein-673b13242](https://www.linkedin.com/in/ross-klein-673b13242/)

## SKILLS AND TECHNOLOGY

**Languages:** Java, Python, SQL, R, C, JavaScript, Haskell

**Libraries/Frameworks:** OpenAI API, Anthropic API, Azure, FastAPI, Postgres, LangChain, LangSmith, Datadog, OTEL, MCP, RAG, Claude code, Codex, Anthropic Skills API, Azure container apps, Azure function apps, React, NodeJS, Harness Dev

## EXPERIENCE

**AGENTIC PRIVATE MARKETS** | **AI Engineer | Forward Deployed Engineer**, New York, NY — **July 2025 – Present**

- Designed production agents with actively managed history, tool calling, MCP tools, knowledge, for various tasks.
- Notable Agents –
  - **Project Chat Agent:** Designed the primary landing-page assistant, providing a unified interface for interacting with projects, search, and executing platform workflows.
    - Design leader for this project, prototyped and integrated into production.
    - OpenAI, GPT-5.6 Terra low reasoning, tool calls, mcp tools, RAG
  - **Artifacts Agent:** Designed a multi-agent workflow that automatically generates PowerPoint presentations.
    - Designed the history management and the last transformation stage (uses function calls to produce pptx) and wrote prompts.
    - OpenAI, GPT-5.6 Terra low reasoning, tool calls, mcp tools, RAG
  - **Sheets Agent:** Designed an AI spreadsheet assistant that researches, formats, and analyzes tabular data.
    - Design leader for this project, prototyped and integrated into production.
    - Anthropic, opus-4.8, skills API, tool calls, mcp tools, RAG
- **Lead Datadog observability integration** using open telemetry traces across the codebase.
- **Ran Langsmith evaluations** to help with model selection, tool, and prompt tuning. Created and cleaned datasets. Focused on Accuracy, Hallucination, Source Conflict, and Recency scores using LLM as a Judge.
- Attended client meetings and product walkthroughs with the founder.
- Worked on backend development, bug fixes, optimizations, cost optimizations.

**ALVAREZ & MARSAL** | **A.I. Team Internship**, New York, NY — **Summer 2024**

- Built follow-up question generator leveraging chat history and retrieved documents and contributed to document extraction and summarization workflows.

**BEAM** | **Teaching Assistant**, New York, NY — **Summer 2023**

- Assisted math teachers in teaching logic and Python programming to accelerated middle school students.

## EDUCATION

**UNIVERSITY OF WISCONSIN-MADISON**, School of Computer & Data Sciences — **May 2025**

*Bachelor of Science in Computer Science*

**Relevant Coursework:** Intro to Big Data Systems, Theory & Design of Programming Languages, Intro to Programming Languages and Compilers, Intro to Algorithms, Machine Organization and Programming, Introduction to Computer Engineering, Intro to Software Security, Data Science Modeling, Linear Algebra, Discrete Mathematics, Calculus I-III

**Publications:** "Judgment testing of adverb topicalization in irregular wh-question." The 14th International Conference of Experimental Linguistics (October 2023, ExLing Society – Athens, Greece)

## PROJECTS

**GRAPHICS ENGINE** — Spring 2026

As a hobby, continually developing a graphics engine in Java. Can be used to develop games or other 3D applications.

**BF JIT COMPILER PROJECT, THEORY & DESIGN OF PROGRAMMING LANGUAGES** — Fall 2024

Developed compiler in JavaScript for the toy programming language BF. Ranked 2nd on class leaderboard. (See GitHub for details)

**AMAZON TRUSTED AI CHALLENGE SUBMISSION** — Summer 2024

Created a research-backed proposal on moderation systems for generative pretrained models with team of 4 students and faculty advisor. (See GitHub for details)

## INTERESTS

Golf, Piano, Music Production, Cooking