Founding AI Engineer

Overview

At Upsolve AI, we’re building the AI-native BI platform from first principles. We’ve already built the foundation: a modern embedded analytics platform designed to work seamlessly with LLMs and agentic systems. Now, we’re going all in on building the next generation of agentic data analysis - think “Cursor for data analytics,” integrated directly into any BI and reporting stack and designed to be embedded into our customers’ products.

We’re a Y Combinator-backed (W24) startup with strong funding and support from top-tier investors (General Catalyst, Samsung Next, Soma Capital, and founders of Airtable and Handshake).

The Founding Team

Ka Ling Wu, Co-founder & CEO, previously led product solution development at nPlan, a Series A AI & Construction tech startup, and before that spent four years at Palantir, where she spearheaded HyperAuto and launched Palantir’s Korea operation.

Serguei Balanovich, Co-founder & CTO, spent seven years at Palantir pioneering early Supply Chain and Manufacturing work and productizing it into HyperAuto.

Together, we bring over 15 years of experience in data engineering and analytics, taking Palantir’s HyperAuto from inception to launch, scaling it to more than 50 enterprise customers in under 2.5 years, and having it featured in Palantir’s DPO S-1. We graduated from Y Combinator in Winter 2024.

The Opportunity

This role is a key hire for Upsolve AI today. We have doubled down on building our agentic data analysis platform, and the quality of our agent performance will determine the trajectory of the business.

As our founding AI engineer, you’ll work directly with the CTO to design and ship real-world LLM/agent solutions. You’ll run experiments, evaluate outcomes, and set product direction based on what proves successful. You won’t just write code you’ll be the deciding voice for what we build, how we build it, and why.

This is not a research role. It’s a high-impact product engineering role in forward-deployed style where your work ships into production and is used by customers daily.

Your first 12 months at Upsolve AI will involve...

• Designing, prototyping, and deploying agentic LLM systems into production with real customer workloads.
• Running experiments to explore architectures, frameworks, and approaches - and driving product direction from their results.
• Building evaluation, observability, and improvement pipelines for agent quality and reliability.
• Integrating with our modern BI platform and customer-facing workflows.
• Establishing best practices for LLM/agent development, from tooling to performance monitoring.
• Embedding and collaborating directly with users and customers to validate what “good” looks like in production.
  
  

What You Bring To Upsolve AI

• 7+ years of engineering experience, including deep experience shipping production LLM systems.
• 2+ years track record of building agentic or multi-step reasoning systems that work in complex, real-world environments.
• Strong command of modern AI/LLM tooling (LangChain, Mastra, DSPy, vector DBs, eval frameworks, observability, OpenAI/Anthropic APIs, fine-tuning).
• Strong engineering ability (TypeScript preferred).
• Experience with databases, data warehouses, data systems, data connections, and big data engineering/analytics.
• Product mindset: ability to make opinionated calls, weigh trade-offs, and push for what delivers real customer value.
• Comfort working in an early-stage environment: high ownership, ambiguity, and rapid iteration.
  
  

What we value and what’s in it for you?

At Upsolve AI, we operate like a high-performance sports team. Hard work is required for success, but when we move together, we don’t just win games, we win championships. We despise bureaucracy and redundant process. Everyone here is a builder who rolls up their sleeves. We value ownership and trust: the person closest to the problem has the authority to solve it. We believe each team member brings a unique strength, and we shape our work to amplify those spikes.

In This Role, You Will Get