Staff AI/ML Engineer, Discovery

About us

Harmonic is the startup discovery engine. We believe in innovation, entrepreneurs, and efficient capital markets. It pains us to see great startup opportunities consistently go undiscovered. So, we dedicated ourselves to mapping out the startup landscape and building the index and tools that ensure the most promising founders get found and funded.

The world's largest and most prolific venture capital firms rely on us to find and invest in the next Google, AirBnB, Uber, Stripe, and Anduril. We play a crucial part in ensuring hundreds of billions of dollars get routed efficiently and that the innovations the world would most benefit from materialize. 

We’ve been growing steadily and need passionate driven people to help us get to the next level. If you want to work in a high-growth environment with some of the most talented folks on the planet and resonate with our values, read on.

About the Discovery team

The discovery team is focused on making it delightfully easy for users to find and understand the startups they’re looking for. We source, transform and index terabytes of data, build models to make sense of the data, and are working at the forefront of applied AI to build magical product experiences. These include powerful structured search interfaces, as well as agentic graphs that fuel natural language search and refinement, along with company research that blends a user’s proprietary data along with Harmonic’s proprietary data and real-time agentic research.

The role

What You'll Do

• Lead our agent and model Infrastructure: Develop robust evaluation frameworks, implement continuous fine-tuning pipelines for our agent graph models, and expand our capabilities across the platform for executing complex workflows autonomously
• Architect advanced search systems: Create and optimize indexing, retrieval and ranking strategies to maximize the effectiveness of natural language search across our data
• Refine alternative discovery features: Develop and refine similarity algorithms, relevance scoring, and intelligent tagging systems that power our search engine
• Make data even more useful: Apply your expertise and creativity to transform our rich datasets into actionable insights and powerful discovery features
• Shape product direction: Collaborate with cross-functional teams to define goals, establish milestones, and execute plans with minimal guidance

Who you are

• A builder.  You’re able to create leverage by educating and empowering other team members, but you get the most joy out of building
• Obsessive. You view the product, no matter how good, as a base that can be continuously improved.
• Low ego. You work in service of making our customers successful.
• A first principles thinker. You are able to get to the crux of user problems, and build solutions with high leverage.
• Endlessly curious. Any new problem is just a new problem, and it doesn't matter that you're not an expert (yet).

What You'll Bring

• Proven experience: 7+ years in Data Science/Engineering or 5+ years industry experience with a PhD.  Demonstrated ability to deliver major projects and features under tight timelines
• Technical expertise: Fluency in Python and SQL with a strong foundation in software engineering principles
• Leadership skills: Ability to architect solutions, provide technical guidance, and rapidly prototype to deliver results
• Search expertise: Deep knowledge of retrieval systems, ranking algorithms, and RAG strategies
• LLM mastery: Expert-level understanding of prompt engineering, LLM evaluation methodologies, and fine-tuning techniques
• Technical toolkit: Experience with backend libraries and tools such as dbt, Dagster, LangChain, and Hugging Face; familiarity with web servers and comfort handling large datasets
• In-office collaboration: Ability to work from our NYC office at least 4 days per week

Compensation:

$185K - $250K Salary + Equity (flexible depending on the level)

Our stack

The Process

Here’s our interview process:

• (30 mins) Introduction Chat: Let's discuss whether Harmonic is the right place for you. We'll review your experience and allocate time for any questions you may have.
• (30 mins) Initial Technical Screen: A technical discussion where we deep dive into how you'd approach a few different problems. Don’t worry, this won’t be a LeetCode-style interview :)
• (~5 hours) Take-Home Exercise: We will send you a description of a challenge akin to ones you would face at Harmonic and ask you to come up with a creative solution. Understanding that schedules can be tight, you’ll have a 48-hour window to complete this at your convenience.
• Interview Day: To streamline the process and respect your time, we try to consolidate the following into one day:
  • (45 minutes each) Team interview: Meet with a member of our Product team to see how you would collaborate and Sang, our Head of Engineering, to make sure there is mutual culture fit.
  • (2.5 hours) Engineering Jam Session: Detailed discussion of the take-home to measure technical skills. This will involve coding with 2-3 engineers on our team.
• (45 min) Get to know Max, our CEO!
• (Async) Reference Checks.

Benefits

• 🩺 Top of the line health, dental and vision insurance, with 100% premium covered
• 📈 401k matching
• 🍜 Free lunch in office
• 🍣 Monthly team dinner (we have a lot of foodies) for each office
• 🚂 Commuter benefits