Applied AI Engineer

Location: New York City (on-site, relocation supported)

Compensation: USD 170,000 – 230,000 + Equity

Hiring: Up to 2 engineers

Company: Confidential (represented by Worthland)

Our client is a well-funded, AI-native startup building a real production marketplace in a space that has historically relied on manual workflows and human judgment.

The company has achieved clear product–market fit , is growing rapidly with a lean team, and uses AI not as an add-on but as the core decision engine of the business. Their systems directly impact matching quality, speed, and operational efficiency at scale.

This is a hands-on role for builders who want to ship AI systems that operate in messy, real-world conditions and materially move business metrics.

As an Applied AI Engineer, you will own end-to-end AI systems powering the marketplace: scoring, matching, ranking, recommendations, and internal automation.

You will work directly with the founder and operations team. There is no separate ML platform team— you own the full loop , from problem definition through production iteration.

This is not a research role and not an infra-only ML position. It is a product-focused, applied AI builder role .

• ~60% Applied AI / ML: model selection, prompting, fine-tuning, evaluation, experimentation
• ~40% Product & Engineering: backend services, APIs, data pipelines, product integration

Week 1

• Develop a deep understanding of existing AI systems (scoring, matching, automation)
• Identify where models perform well vs. where they fail in production
• Partner with the ops team to understand manual overrides and edge cases
• Align with the founder on the highest-leverage problems to solve first

Weeks 2–4

• Ship meaningful improvements to scoring/matching models and related product features
• Begin reducing manual ops load through smarter automation
• Take ownership of the AI roadmap, using the founder as a thought partner—not a project manager

• Build and improve matching, scoring, and ranking systems that directly affect marketplace outcomes
• Design explainable and calibratable AI outputs so users can understand and trust decisions
• Compare new inputs against historical outcomes to surface meaningful signal
• Iterate based on real-world feedback, not just offline metrics

• Replace manual review steps with intelligent AI-driven workflows
• Build feedback loops where human overrides improve model performance over time
• Identify and surface stuck or anomalous cases requiring human attention

• Generate structured summaries from unstructured data (documents, notes, transcripts)
• Build retrieval-based assistants and role-specific AI tools
• Ship features that are user-facing, measurable, and tightly integrated into the product

• Design evaluation frameworks that reflect real business impact, not just model accuracy
• Implement experimentation and A/B testing to validate improvements
• Make pragmatic tradeoffs (prompt vs. fine-tune, off-the-shelf vs. custom)

• 4+ years of engineering experience with AI/LLM features shipped to production
• Experience owning AI systems end-to-end in a real product environment
• Hands-on experience with LLM APIs, embeddings, vector databases, and evaluation
• Strong backend or full-stack engineering fundamentals
• Comfort working with imperfect data and evolving requirements

• Experience at an AI-native startup or as a founding / early engineer
• Background in recommendation systems, matching, search, or automation
• Product intuition and comfort making fast, high-impact decisions
• Willingness to be hands-on rather than operating through layers of abstraction

• Work only on ML platforms or infrastructure without owning product features
• Focus on research, publishing, or offline experimentation
• Require clean datasets and mature infra before shipping
• Have only built basic RAG demos as your main AI experience
• Prefer large, highly structured organizations

• High ownership and direct impact on core business metrics
• Close collaboration with the founder and decision-makers
• AI-first product where models are central, not decorative
• Competitive compensation, meaningful equity, and long-term upside

• Intro Call (30 minutes): High-level screening to assess role fit, hands-on AI experience, and mutual interest in the opportunity and company stage.
• Technical Interview (60 minutes): Applied system design discussion focused on a real-world AI problem (e.g. matching, ranking, automation), evaluating practical decision-making and tradeoffs.
• Behavioral Deep Dive (60 minutes): Review of past experiences, ownership mindset, and working style in high-autonomy, fast-moving environments.
• On-Site (Full Day): Collaborative working session solving real problems with the team to assess technical depth, problem-solving approach, and collaboration in practice.