Applied AI/ML Engineer

About the Company

We are an AI-focused career and recruitment platform dedicated to building behavioral intelligence tools for the future of hiring. Our goal is to develop an intelligent system that goes beyond resumes to aid organizations in making more insightful hiring decisions. Through AI-powered career coaching, we assist candidates in preparing comprehensively for their interview processes. Simultaneously, our AI-driven recruitment technology enables companies to source, evaluate, and hire candidates more effectively and efficiently.

Working with us means joining a dynamic, fast-paced team that values speed, accountability, and AI-first execution to produce outstanding products. You will collaborate directly with the founders and CTO, gaining substantial influence on both the product and company during the early growth phase.

Role Overview

As an Applied AI/ML Engineer, you will create the initial version of Hiring Intelligence models—systems that convert recorded interview transcripts into evidence-based candidate profiles and explainable ranked shortlists that recruiters rely on.

This position emphasizes applied machine learning, focusing on Large Language Model (LLM)-based structured extraction supported by engineering discipline, transparent scoring models, and rigorous evaluation to validate system effectiveness. You will be responsible for the end-to-end pipeline from raw transcripts to structured, queryable data, culminating in production deployment.

Our core principle is that LLMs extract evidence only—they do not make final predictions. Every score produced must be decomposable into verifiable evidence for recruiters or auditors. If this transparent, evidence-based modeling excites you, this role is an excellent match.

Key Responsibilities

• Develop an extraction pipeline using LLM-based structured extraction to capture competency evidence from interview transcripts. This includes managing versioned prompts and evaluation rubrics, maintaining output provenance, capturing evidence quotes, and measuring accuracy against expert labels with an engineering mindset.
• Create a first iteration scoring and matching model that transparently ranks candidates based on weightings of role requirements. Interpretability is mandatory, and this groundwork will support future learning-to-rank approaches as recruiter preference data grows.
• Design and maintain an evaluation framework that measures precision at K against recruiter choices, agreement with expert annotations, proper calibration, and subgroup fairness checks. Each model version will be accompanied by a thorough evaluation report.
• Manage label operations by coordinating expert labeling workflows with subject matter experts and collecting recruiter preference data, alongside overseeing training data snapshots and model version control.
• Own the full product lifecycle—from prototyping through evaluation, deployment, and monitoring—using technologies such as Python, PostgreSQL, AWS S3, and standard batch and queue processing infrastructure.
• Work independently on ambiguous challenges by analyzing problems, proposing written solutions, and driving projects to production with minimal supervision.

Essential Qualifications

• Between 4 to 6 years of practical experience developing and deploying ML/AI products in production environments, with responsibility for ongoing system reliability—not just research or experimental models.
• Expertise in using LLM components responsibly: building structured extraction or LLM pipeline systems with versioned prompts and measured accuracy, treating LLM results as evaluable data sources rather than absolute truths.
• Strong foundation in classical machine learning techniques, including linear and gradient boosting models, regularization, and rigorous train/test protocols. Understanding of overfitting risks especially with limited labeled data.
• Proficiency in experimental evaluation design, able to critically assess metrics like precision@K, AUC, and calibration. Confidence to declare models ineffective if such evidence arises.
• Advanced skills in Python programming and familiarity with contemporary ML frameworks such as PyTorch, scikit-learn, and Hugging Face. Comfortable working with data engineering tools including SQL, batch processing, and containerization (Docker).
• A builder mentality with proven ability to take ownership and independently deliver end-to-end production-grade products.

Additional Advantageous Skills

• Experience in learning-to-rank or recommender system development.
• Background or awareness in hiring or assessment technology domains, including related regulatory frameworks such as NYC LL144 or the EU AI Act.
• Competence in fairness and bias measurement pertaining to ranked or scored outputs.
• Familiarity with speech recognition, audio/video feature extraction technologies, anticipated in future product roadmaps.
• Knowledge of vector search techniques and embedding-based retrieval.
• Previous work experience in early-stage startup settings or founder-led teams.

Role Clarifications

• This is not a research-heavy position: no academic paper writing or novel model design required.
• It is not a prompt-engineering-only role; responsibilities include pipelines, scoring models, and comprehensive evaluation.
• Not a notebook-only data science role; all developments are expected to be productionized.

Success Metrics

• Within 30 days: Operational extraction pipeline generating versioned, evidence-linked outputs on authentic interview transcripts, alongside an active evaluation framework.
• Within 60 days: Launch of the initial scoring and ranking model running in shadow mode, initiation of expert-labeling workflow, and generation of preliminary accuracy and fairness assessments.
• Within 90 days: Delivery of a statistically rigorous analysis of model quality compared to recruiter selections, with clear recommendations on components to retain, improve, or abandon.

What You Will Receive

• Unrestricted access to the latest AI tools including Claude and coding assistants, along with comprehensive infrastructure support.
• Direct collaboration opportunities with CTO and founders, influencing both technical direction and product development.
• Ownership over a pivotal AI intelligence layer that shapes real-world hiring decisions.