Senior Agentic Platform Engineer II

Porch Group is a leading vertical software and insurance platform and is positioned to be the best partner to help homebuyers move, maintain, and fully protect their homes. We offer differentiated products and services, with homeowners insurance at the center of this relationship. We differentiate and look to win in the massive and growing homeowners insurance opportunity by

1) providing the best services for homebuyers

2) led by advantaged underwriting in insurance

3) to protect the whole home

As a leader in the home services software-as-a-service (“SaaS”) space, we’ve built deep relationships with approximately 30 thousand companies that are key to the home-buying transaction, such as home inspectors, mortgage companies, and title companies.

In 2020, Porch Group rang the Nasdaq bell and began trading under the ticker symbol PRCH. We are looking to build a truly great company and are JUST GETTING STARTED.

Job Title: Senior Agentic Platform Engineer II

Location: India

Workplace Type: Remote

Job Summary

The future is bright for the Porch Group , and we’d love for you to be a part of it as our Senior Agentic Platform Engineer II.

As a Senior Agentic Platform Engineer II , your charter is to build out theAgentic Platform at Porch Group — the shared foundation every team builds AI on. You will own how we build retrieval pipelines, deploy and manage remote agents, create MCP servers, route and govern model access, and evaluate all of it. When a new use case arrives, this team builds the scaffolding and often a proof of concept, then hands off to the feature team that owns the production system.

AI adoption here has moved fast: agentic coding tools are in active rollout across multiple business units, autonomous agents already open merge requests against production repositories, and a centralized LLM gateway is live.

The platform runs primarily on Python with some TypeScript /Node, on Kubernetes and Google Cloud , alongside vector stores for retrieval, LLM evaluation and observability tooling, a workflow orchestration platform for production AI pipelines, and a growing fleet of internal MCP servers.

Porch's product services are predominantly JVM-based ( Java, Kotlin, Scala ), and the agents you build will operate on that code.

What You Will Do As A Senior Agentic Platform Engineer II

Build out the Agentic Platform as a product for internal teams — find what teams keep rebuilding and turn it into a paved path with sane defaults.

• Design, build, and operate internal MCP servers, and own the standards for them: auth (OAuth 2.1), tool and schema design, versioning, and testing — plus the scaffolding, templates, and registry that make creating one a solved problem.

• Build the platform layer for remote agents: how an agent gets deployed, scheduled, credentialed, observed, rate-limited, and safely retired, across conversational, scheduled, and event-triggered shapes.

• Build the guardrails that make autonomous agents trustworthy — scoped permissions, human-in-the-loop approval gates, sandboxed execution, audit logging, spend controls, and hard blast-radius limits.

• Serve as technical lead for retrieval-augmented generation — shared patterns for ingestion, chunking, embedding, and retrieval; vector stores; incremental sync; and retrieval quality owned as a measured discipline rather than a guess.

• Own evaluation and observability for AI systems — eval harnesses, golden datasets built from the questions people actually ask and calibrated against domain experts, prompt versioning, and instrumentation through Opik, Datadog, and OpenTelemetry.

• Advance the agentic developer experience — the internal plugin and skills library encoding Porch conventions, our code review agents, the autonomous bug and security remediation agents, and tooling that makes fleet-wide migration across dozens of JVM repositories tractable.

• Extend and operate the LLM gateway (Portkey) — provider and model coverage, routing, secure access for business units outside the core Agentic Platform, and governance including access controls, PII monitoring, and cost attribution.

• Build reference implementations and proofs of concept that establish a pattern, then hand off cleanly to the team that will own it in production.

• Set standards and playbooks, mentor engineers across business units adopting AI tooling, give leadership straight answers on build versus buy, and participate in an on-call rotation for Agentic Platform services.

What You Will Bring As A Senior Agentic Platform Engineer II

Bachelor’s degree in Computer Science, Engineering, or related field (Master’s preferred)

8+ years of professional software engineering experience, including 2+ years building LLM-powered systems that reached production and at least 2 years in a senior or technical lead role.

• Languages: production-grade Python is the primary language you will write. JVM fluency required — you can read and confidently modify a large Java or Kotlin codebase and its build tooling (SBT, Maven, Gradle), since the agents you build operate on that code; Scala is a plus, not a requirement. TypeScript/Node.js strongly preferred.

• Hands-on MCP experience — building servers or clients, tool and schema design, auth — and agent orchestration with LangGraph or equivalent, including context and state management.

• RAG in production — vector stores such as PostgreSQL/pgvector, embedding models, chunking and retrieval strategies, incremental sync, and how to tell whether retrieval is actually working.

• Experience evaluating non-deterministic systems — eval harnesses, regression suites, quality gates, prompt versioning — and rigorous reasoning about latency and cost.

• A working practitioner of agentic coding tools. Real production work with Claude Code or comparable, and informed opinions about where agents fail.

• Security judgment for agentic systems — prompt injection defense, sandboxed execution, scoped credentials, audit logging, and containing an agent's blast radius.

• Kubernetes and cloud experience, ideally GCP (GKE, IAM, GCS, Cloud SQL, Pub/Sub, BigQuery), plus GitLab CI/CD, Terraform, Docker, and observability tooling (Datadog, Prometheus, OpenTelemetry).

• A platform builder's instincts and strong written communication — turning a business problem into a reusable capability, proving it with a focused proof of concept, and writing standards and clean, tested code others can build on.

• Comfort participating in an on-call rotation, ability to overlap 2-3 hours with Pacific Standard Time (PST) core US business hours, and a collaborative approach in a trustful and respectful environment.

• Proven experience working with US based business teams.

• Excellent written and communication skills in English language.

Must currently reside in or are located within 150 km of Greater Delhi NCR, Greater Bangalore, or Greater Pune.

Working Hours: 8 Hours (Excluding breaks) — 4 Core US Hours Overlap (7:30pm–11:30pm IST), 4 flexible hours (IST)

Workspace: A quiet space to work, an internet connection of at least 30 Mbps download | 10 Mbps upload

Nice to Have

Building developer platforms or internal tooling with real adoption metrics behind it.