AI Prompt Engineer

Location: New York, NY | Type: Full-time

About Voiceops

We are unlocking the richest dataset in the enterprise: human conversation.

Companies interact with customers millions of times a day. This data is dense, high-fidelity, and historically unusable. Voiceops changes that. We turn massive streams of audio and text into structured, queryable data that powers decision-making across the entire business.

We serve multiple $1B+ revenue companies, including Fortune 500 leaders, who rely on us to automate QA, score leads, and feed live market data into their systems.

The momentum:

• Our software works. Our customers use Voiceops today to automate QA on 100% of their calls, populate knowledge bases with real-time data, source real-time market information, and enrich their CRM.

• We are backed by the best. We raised a $5M Seed round co-led by Bonfire Ventures and Twelve Below, with support from Village Global, Precursor, Not Boring Capital, and executives at Anthropic and Airbnb.

• We are building the future. The shift to agentic software is a 10x larger opportunity than the SaaS era, and agents are starved for data and context. We are feeding these hungry machines with high-fidelity context (conversation data) that enables the next generation of software to operate autonomously.

About the Role

We are looking for an AI Prompt Engineer to shape how our artificial intelligence thinks, reasons, and understands the world.

This is not a software engineering role. You will not be writing code. Your work will happen in natural language. You will design the instructions and reasoning frameworks that allow our AI to understand messy human conversations, interpret intent, and transform ambiguity into structured meaning.

This role requires both creativity and discipline. You need the imagination to design how an AI should approach vague or contradictory information, and the rigor to test, measure, and refine that thinking at scale.

The Challenge

We do not build one off models for each customer. We build general purpose reasoning systems. Your work will include:

Collaborating with users. When a user gives feedback such as "This model is not working," your prompts must help the AI audit the examples, identify inconsistencies, and propose a clear path forward.

Detecting patterns. You will design reasoning chains that allow the AI to scan thousands of conversations and discover call types, trends, and structures without manual labeling.

Reasoning economically. You will ensure the AI delivers high quality reasoning using fast and cost efficient models while keeping latency low.

What You Will Do

Design meta prompts. Create natural language instructions that guide the AI in building and improving models based on user needs.

Create reasoning chains. Break down complex cognitive tasks into clear and reliable steps that the AI can follow consistently.

Infer structure from raw data. Build prompts that identify fields, categories, and relationships within unstructured conversations.

Evaluate and improve. Read model outputs to judge clarity and correctness, use evaluation sets to measure performance, analyze failures, track experiments, and refine the system through iteration.

Optimize reasoning. Design workflows that maintain high quality while keeping cost and latency low.

About You

We do not care about your degree. We care about your ability to think clearly, write precisely, and solve unstructured problems. You may come from a liberal arts education, a research background, a customer facing role, or you may be entirely self taught. What matters is your ability to analyze ambiguity, communicate with clarity, and design well structured reasoning.

We are looking for:

Exceptional command of language. You understand nuance, tone, and ambiguity. You can write instructions so clearly that they cannot be misunderstood.

Systems thinking. You can visualize how prompts shape data, how that data becomes structured, and how that structure influences examples, evaluations, and downstream systems.

Scientific rigor. You form hypotheses, run experiments, record results, and improve through iteration.

Technical fluency. You do not need to write code, but you are comfortable with concepts such as JSON, logic flows, and data structures, and you are willing to learn the tools of the trade.

We are looking for "Magical Teammates" who:

• Move Fast, Learn Fast: You measure progress in hours and days, not weeks. You prioritize speed of execution and learning over perfect information.

• Drive to "Done-Done": You ensure the whole job is complete so no one else has to think about it ever again.

• Jump All The Way In: You spot the work nobody explicitly assigned and you just do it. You consistently shoulder an amount of responsibility that wows people.

• Work Out In The Open: You hate silos. You pull people into your work, hunt for feedback, and keep the team moving forward together.

Why Join Voiceops?

• High Expectations, High Reward: We aren't a "9-to-5, rest and vest" shop. We are a high-performance team. If you perform, we pile on responsibility and ownership (equity + responsibility) regardless of your tenure.

• Creative Destruction: Our team has a decade of experience in this space, but we are constantly learning. We iterate, we adjust on the fly, and we aren't afraid to throw away logic to get to a simpler, more powerful answer.

• Growth by Subtraction: We believe complexity is the enemy. We strive to strip things down to their essence.

• Define a New Discipline: "Prompt Engineering" is still being defined. You will be at the forefront, establishing the best practices for how humans direct AI attention.

Benefits

• Competitive Salary & Significant Equity

• Top-tier Medical, Dental, and Vision

• Unlimited PTO

• Office in the heart of NYC (Union Square)

• The hardware and tools you need to do your best work

Compensation

• Salary: $90,000 – $115,000 USD

• Equity: Significant early-stage equity package (0.1% - 0.25%)