Cloud Architect with AWS AgentCore

We’re looking for a hands-on AWS AgentCore Cloud Architect to independently design, build, and deploy AgentCore solutions on AWS—integrating use cases, optimizing workflows, and ensuring secure, scalable execution.

Key Responsibilities

• Design and build AgentCore workflows and automation on AWS (Lambda, Step Functions, S3, DynamoDB, API Gateway, ECS/EKS, Bedrock).
• Integrate AgentCore with external systems, APIs, and data sources.
• Optimize cloud architecture for performance, cost, and security.
• Troubleshoot AgentCore orchestration and AWS environment issues.
• Build reusable modules, templates, and scripts to accelerate deployments.
• Partner with architects, product, and engineering to deliver end-to-end use cases.
• Create technical documentation and deployment guides.

Required Skills

• Hands-on experience building and deploying agents with the AWS AgentCore framework (workflows, triggers, orchestration) and core AWS services (AI, compute, storage, networking, IAM, serverless).
• Experience with event-driven architectures and API integrations.
• Proficiency in Python and/or Node.js for automation and scripting.
• Infrastructure as Code using Terraform and/or CloudFormation (preferred).
• Strong debugging, problem-solving, and performance tuning skills.
• CI/CD experience (Git, CodePipeline, Jenkins, etc.).

Industry Experience (1 )

• Life Sciences R&D: R&D/lab/clinical data flows; regulated workloads (GxP/21 CFR Part 11); auditability, lineage, and access controls.
• Supply Chain: Planning/logistics/inventory use cases; event-driven integration across ERP/WMS/TMS; real-time visibility.
• Financial Services: Secure, compliant cloud delivery (PCI/SOC2); strong IAM, encryption/KMS, observability, and HA.

Consultative-Ability Scorecard (AWS AgentCore Cloud Architect)

Purpose: Use this scorecard to consistently evaluate consultative ability for a hands-on AWS AgentCore Cloud Architect: discovery, stakeholder alignment, structured problem-solving, executive communication, and decision facilitation—grounded in technical realities (security, cost, delivery feasibility).

Scoring scale (1–4)

• 4 – Strong: Leads the room; drives clarity and decisions; communicates crisply to both exec and engineering audiences; produces reusable artifacts.
• 3 – Solid: Runs discovery well and communicates clearly; occasional gaps in structure or executive-level framing; needs minor guidance.
• 2 – Developing: Understands consulting motions but struggles to structure ambiguity, tailor to audience, or drive decisions; artifacts lack clarity.
• 1 – Weak: Reactive; unclear communication; cannot translate business needs into actionable technical direction; limited stakeholder influence.

Weighted criteria (total = 100 points)

#

Dimension

Weight

What “good” looks like (observable behaviors)

Evidence to capture

1

Problem framing & hypothesis thinking

15

Quickly clarifies goals, constraints, and success metrics; proposes hypotheses and validates with stakeholders; separates symptoms from root causes.

2-minute problem statement; explicit assumptions; stated success metrics.

2

Discovery & requirements (use cases, workflows, edge cases)

15

Runs structured discovery; maps current vs future workflow; identifies actors, triggers, inputs/outputs, failure modes; documents decisions.

Workshop approach; sample questions; workflow/process map; edge cases list.

3

Stakeholder management & alignment

15

Identifies decision-makers; manages conflicting priorities (security vs speed vs cost); creates shared understanding; drives closure.

Example of disagreement resolved; RACI/decision log; alignment plan.

4

Executive communication (clarity, concision, story)

15

Communicates outcomes, tradeoffs, and risks in business language; produces crisp exec summaries; avoids jargon overload.

Verbal exec summary; headline-first narrative; crisp next steps.

5

Technical translation (business ↔ architecture)

15

Converts goals into architecture decisions (integration, IAM, observability, cost); ties design choices to business value and constraints.

Options rationale; NFR mapping; security/cost implications.

6

Decision facilitation & tradeoff management

10

Presents options with pros/cons; quantifies cost/effort/risk; recommends; secures decisions and documents them.

Options matrix; decision note/ADR; risks & mitigations.

7

Delivery orientation (plan, milestones, dependencies)

10

Produces pragmatic phased plan; identifies dependencies/access needs; proposes MVP; anticipates operating model requirements.

Phased roadmap; dependency/access checklist; MVP definition.

8

Written artifacts quality (client-ready outputs)

5

Produces clean, structured documentation; easy to follow; reusable templates/modules guidance.

Sample doc outline; runbook; deployment guide.

Suggested interview structure (45–60 minutes)

1. Case prompt (15 min): Design an AgentCore-based workflow that integrates with external systems and data sources; include security (IAM/KMS), observability, and cost controls.
2. Deep dive (15 min): A past engagement from discovery → options → decision → delivery; focus on stakeholder alignment and tradeoffs.
3. Executive summary (5 min): Candidate delivers a 2-minute exec readout (headline, recommendation, risks, next steps).
4. Written exercise (10 min): Draft a short “Decision & Next Steps” note suitable for email/Confluence.

Standard prompts & questions (use at least 6)

• Problem framing: “What’s the objective, and how will we measure success in 30/60/90 days?”
• Assumptions: “What assumptions are you making? Which ones are riskiest and how would you validate them?”
• Discovery: “Walk me through your discovery plan for week 1—who do you meet and what outputs do you produce?”
• Failure modes: “What failure modes do you plan for in agent workflows (timeouts, bad inputs, permission failures, non-determinism)?”
• Stakeholder alignment: “Tell me about a time security/compliance blocked the initial design. How did you adapt and still deliver?”
• Tradeoffs: “Give 2–3 architecture options and recommend one. What do we gain/lose (cost, time-to-value, risk)?”
• Executive communication: “Explain your recommendation to a VP in 90 seconds—define acronyms as needed.”
• Delivery: “What does your MVP look like? What do you defer, and how do you prevent it from becoming tech debt?”

Decision rubric

• Hire / Strong Yes: Weighted score ≥ 80, and no dimension below 3 in criteria #1–#6.
• Hire / Yes: Weighted score 70–79, with at most one dimension below 3 in criteria #1–#6, plus a clear improvement plan.
• Leaning No: Weighted score 60–69 or two core dimensions (#1–#6) below 3.
• No Hire: Weighted score < 60 or any critical failure in executive clarity, alignment, or decision-making.

Interviewer notes

• Key quotes (verbatim):
• Strengths observed:
• Risks / gaps:
• Follow-ups to validate: