Forward Deployed Engineer - Agentic AI

Job Description

Multiple AI roles_FDE–Agentic AI/AI Prompt Engineer/AI Lead Engineering/AI Architect / FDE – Healthcare_New York, Bay Area, Dallas, or Chicago

\n

\n

Locations: New York, Bay Area, Dallas, or Chicago

\n

Work model: hybrid, with no fully remote option.

\n

Travel: approximately 25% when necessary.

\n

Office attendance: the exact weekly onsite schedule remains pending.

\n

Compensation guidance: USD 150,000–200,000 annual base, with potential flexibility.

\n

\n

Full JDs:

\n

JD1:Forward Deployed Engineer – Agentic AI

\n

\n

This role combines senior client engagement with hands-on engineering, translating business challenges into working AI solutions and taking them from rapid proofs of value toward MVP and production delivery.

\n

\n

Required background:

\n

\n

- 8+ years in full-stack engineering, solution architecture, or technical consulting, with significant client-facing delivery.

\n

- Strong Python plus TypeScript, Java, or Go.

\n

- Practical experience with LLMs, agent orchestration, tool calling, planning/execution loops, and memory patterns.

\n

- RAG, embeddings, chunking, vector databases, and retrieval evaluation.

\n

- APIs, data pipelines, enterprise integrations, and sound software architecture.

\n

- Cloud-native engineering on AWS, Azure, or GCP, including containers, CI/CD, and observability.

\n

- Ability to explain technical decisions and measurable business value to senior stakeholders.

\n

\n

Prioritize engineers who personally build solutions, develop evaluation methods, and address security, access controls, human oversight, reliability, latency, and cost. Production GenAI experience and integrations with platforms such as ServiceNow, Salesforce, SAP, Microsoft, Databricks, or Snowflake are preferred.

\n

----------

\n

JD2: AI Prompt Engineer / Agentic AI Engineer

\n

\n

This is a hands-on engineering role focused on developing agents that reason, use tools, and execute enterprise workflows.

\n

\n

Key requirements:

\n

\n

- Experience building LLM applications and agent workflows using Semantic Kernel, LangChain, LangGraph, AutoGen, CrewAI, or comparable frameworks.

\n

- Hands-on MCP servers, tools, API orchestration, and enterprise connectivity.

\n

- Python, TypeScript, or similar programming languages.

\n

- Prompt optimization, agent planning, memory, context management, and multi-step execution.

\n

- RAG, embeddings, vector databases, and knowledge grounding.

\n

- Familiarity with Azure AI Foundry, Azure OpenAI, AWS Bedrock, Google Vertex AI, or comparable platforms.

\n

- Agent evaluation, testing, deployment, monitoring, troubleshooting, security, and governance.

\n

\n

Enterprise copilots, AI assistants, and healthcare or regulated-system integrations are preferred. The supplied JD does not specify a minimum number of years, so please assess demonstrated engineering depth rather than imposing the FDE or platform-engineering experience threshold.

\n

-----------

\n

JD3: AI Lead Engineering – Enterprise AI, Cloud & Platform Transformation

\n

\n

This role combines AI engineering with enterprise platform design, cloud transformation, and production operations.

\n

\n

Required experience:

\n

\n

- 12+ years across AI, DevOps, cloud engineering, SRE, or platform engineering.

\n

- 3+ years of hands-on AI-stack experience.

\n

- 5+ years designing enterprise platforms.

\n

- 3+ years implementing AIOps or observability platforms.

\n

- Enterprise cloud transformation, production AI/ML environments, and mission-critical or regulated platforms.

\n

- Strong coordination with architecture, operations, security, and delivery teams.

\n

\n

Python is mandatory. Relevant technologies include GenAI, LLMs, RAG, AgentCore, MCP, AWS Bedrock, SageMaker, vector databases, model lifecycle management, and AWS/Azure/GCP. Experience with engineering copilots such as Kiro or GitHub Copilot is relevant.

\n

\n

Infrastructure automation, CI/CD, Kubernetes, Docker, service meshes, and tools such as Terraform, Datadog, Dynatrace, Splunk, Prometheus, and OpenTelemetry are useful. Specific tools are listed as preferred in several areas; do not require every named product. The required AIOps/observability experience should still be demonstrated.

\n

------------

\n

JD4: AI Architect / FDE – Healthcare

\n

\n

This role focuses on AI-enabled enterprise architecture, GCP engineering, and integration in a healthcare environment.

\n

\n

Required background:

\n

\n

- 12+ years of overall IT experience, with substantial cloud, enterprise, integration, or AI/ML architecture experience.

\n

- Strong hands-on GCP and cloud-native architecture.

\n

- Vertex AI, GenAI solution design, prompt engineering, model integration, governance, and production deployment considerations.

\n

- Strong Java and Spring Boot, including microservices, REST APIs, security, and scalable backend design.

\n

- Enterprise integration connecting modern and legacy platforms.

\n

- Ability to evaluate and apply AI productivity tools such as Claude and Antigravity.

\n

- Strong consulting and communication skills with senior business and technology stakeholders.

\n

\n

Healthcare, clinical operations, payer/provider integration, regulated environments, and relevant Google Cloud certifications are preferred. GCP/Vertex AI and Java/Spring Boot are central requirements for this search.

\n

\n

Shared employment and compensation guidance

\n

=========

\n

All four opportunities are direct hires with the hiring employer.

\n

\n

- Locations: New York, Bay Area, Dallas, or Chicago.

\n

- Work model: hybrid, with no fully remote option.

\n

- Travel: approximately 25% when necessary.

\n

- Office attendance: the exact weekly onsite schedule remains pending.

\n

- Compensation guidance: USD 150,000–200,000 annual base, with potential flexibility up to USD 220,000 for exceptional candidates, subject to approval.

\n

- Benefits include medical/dental/vision coverage, disability and life insurance, a 401(k) with company match, paid leave, and parental leave. A performance-based bonus is indicated, with individual terms to be confirmed.

\n

\n

Please capture salary expectations without treating the indicated range as an automatic exclusion. Use our approved candidate invitations, which omit the salary range.

\n

\n

US citizens and Green Card holders are preferred. H1B transfers are possible. The client indicated that OPT, STEM OPT, and EAD candidates may be considered where they have completed their master’s degree and started working in the US, subject to individual confirmation.

\n

\n

For H4-EAD or other authorization cases, provide the actual category, expiration date, and sponsorship/transfer requirements. No minimum two-year EAD validity requirement has been confirmed. Candidates must be willing to join the hiring employer directly rather than BestInfo’s W-2 payroll.

\n

\n

Required deliverables:

\n

\n

Detailed questionnaires are not required for initial submission. Reuse existing information, reconfirm anything that has changed, and collect only missing essentials.

\n

\n

For each qualified and interested candidate, provide:

\n

\n

1. An updated resume/profile in Word format.

\n

2. The completed summary table below.

\n

3. A brief role-fit note identifying the proposed role and strongest evidence of relevant engineering experience.

\n

4. Confirmed interest and interview availability, including time zone.

\n

\n

Please also include contact details, current employment status/title/company, work authorization and sponsorship needs, travel acceptance, and the role-specific skills needed to establish fit.

\n

\n

Only request a brief technical example if the existing profile does not clearly demonstrate the candidate’s personal contribution. Two or three sentences covering the solution, technologies, production status, and results are sufficient.

\n

For role , capture the required AI, platform-design, and AIOps/observability experience separately.

\n

\n

Assessment and interview guidance:

\n

\n

For selected candidates, the evaluation includes a 45-minute CoderPad assessment covering cloud expertise, technical panel and client interviews, and HR discussions. The sequence follows the instructions for each candidate; CoderPad does not necessarily precede L1 for candidates already in process.

\n

\n

Candidates can familiarize themselves with the platform at This is the general website, not their assessment link. At the assessment stage, coordinate availability beforehand and explain that the test must be completed within 24 hours of receiving the individual link.