I AM HIRED

Неделя 310 ₽Месяц 630 ₽3 месяца 1 680 ₽Год 5 680 ₽

Tp 9Sc6eSKv — New York City Metropolitan Area

ML и AI, AI-инженерия · вчера, 21:32 · 18.11.2026

Зарплата не указана

Откликнуться

Компания: ••••••••

Локация: New York City Metropolitan Area

Формат: не указан

Занятость: Полная занятость

Грейд: не определён

About 9MtG5

9MtG5 is building the intelligence layer for critical frontline workforce management.

Frontline workers move between shifts, facilities and locations, and every assignment brings its own qualification and compliance requirements. Today that work runs on manual handoffs across disconnected systems, which slows deployment and drives up cost.

Our NT agents run these workflows across the systems organizations already use:

• Recruiting: sourcing, screening and submitting candidates.

• Onboarding: building requirement checklists, verifying documents and licenses, and following up on outstanding steps.

• Continuous compliance: monitoring credentials and running primary source verification.

We are extending our agents into scheduling and deploying workers across new locations and industries. We are an early-stage team based in New York.

nT ERz6CJwdEj7vwGGk: New York, NY · Type: Full-time

Work authorization: This role is open only to candidates who are already authorized to work in the United States without visa sponsorship, now or in the future. 9MtG5 does not sponsor, transfer or extend work visas of any kind, at any stage of employment. If you will need sponsorship at any point, including when a temporary work permit expires, please do not apply. Applications that misstate work authorization will be removed from consideration.

You will build the agents that do the work: reading documents, screening candidates, writing to workers, deciding what needs a human, and soon, deciding who can work where and when. Your work is measured by how much of a real workflow an agent completes correctly, not by demos.

What you'll do• Build and improve agents across recruiting, onboarding and continuous compliance: matching and screening candidates, building requirement checklists, and classifying, extracting and verifying documents.

• Design extraction pipelines that turn messy inputs (email threads, resumes, licenses, certificates, background-check results) into structured, schema-validated records in our data model.

• Build generator–verifier loops for worker and recruiter communication, so drafts carry the right facts, the right intent and the sender's own voice.

• Give agents memory: capture conversations across email and SMS so they keep context and never re-ask for something already collected.

• Own escalation logic and confidence scoring. Agents recommend and act when they are sure, and route to a person with a clear recommendation when they are not.

• Build and maintain evals: labelled sets of real cases, end-to-end simulations, and shadow runs against live decisions, with divergence targets a system must hit before it replaces the old one.

• Help lay the groundwork for eligibility and scheduling agents that reason over each worker's qualifications, history and preferences against each location's requirements.

• Ship everything as configurable platform capability. A new customer or industry is configuration, not a one-off build.

What we're looking for• 1+ years of software engineering experience, including hands-on work building with LLMs.

• Strong TypeScript or Python. You are comfortable owning a feature from data model to deployed service.

• Hands-on experience with structured outputs, tool use, retrieval and multi-step agent workflows, and a clear view of when not to use an agent.

• An evals-first habit: you build the test set before tuning the prompt, and you can explain why a metric moved.

• Comfort with messy real-world data and documents, including OCR output, inconsistent formats and missing fields.

• Care with sensitive personal data and an instinct for audit trails and human review where mistakes are costly.

• Clear written communication. You can explain an agent's failure mode to an operations lead in plain words.

Nice to have• Experience with the Vercel XV pX8, Zod and durable workflow engines (retries, idempotent steps, human-in-the-loop pauses).

• Background in workforce management, staffing, HR tech, scheduling or compliance software.

• Experience integrating with applicant tracking, vendor management or HR systems, or with browser automation.

• Experience with document NT: classification, layout-aware extraction or verification against a source of record.

Поиск по вакансиям

Название компании скрыто

Названия компаний и их продуктов мы не публикуем на сайте — они доступны в Telegram-боте I AM HIRED. Откройте эту вакансию в боте: увидите название, сможете откликнуться и настроить рассылку подходящих вакансий.

Открыть вакансию в боте