Software Engineering SMTS / LMTS - AI Core (REMOTE)

Salesforce•San Francisco, CA

18h•Remote

About The Position

Salesforce is hiring an engineer for the AI Core team at Qualified, a high-ownership, high-velocity engineering environment backed by Salesforce. The team builds retrieval systems and agent infrastructure that enable AI agents to find, trust, and use customer knowledge, along with developer tooling that helps engineers build production agents safely and quickly. Engineers own work end to end, from scoping and architecture through implementation, testing, deployment, monitoring, and iteration based on customer feedback. The role calls for a technical leader who can turn ambiguous, high-stakes AI challenges into durable systems and raise engineering standards across the team. The engineer will lead initiatives involving RAG and hybrid search, agent frameworks and orchestration, content ingestion and lifecycle controls, transcript processing, and evaluation and developer tooling. They will collaborate with product, design, and engineering leadership to set priorities, make architectural tradeoffs, and measure answer quality. The role includes operational ownership, on-call participation, incident leadership, mentoring, and using agentic development tools while maintaining a high validation bar.

Requirements

Proven experience shipping production AI or backend systems in a high-ownership environment, with recent focus on LLM-powered applications.
Expertise building retrieval systems using RAG and hybrid search, including chunking and indexing strategy, embeddings, vector and lexical search, reranking, and retrieval-quality measurement.
Expertise building developer tooling for production agents, potentially including multi-agent orchestration frameworks, agent and tool registries, and shared tracing and evaluation infrastructure.
Experience owning platform initiatives from design through adoption across areas such as retrieval and indexing, LLM provider integrations, and agent orchestration, and evolving platforms as usage grows.
Demonstrated technical leadership, including leading significant projects, weighing tradeoffs on non-trivial problems, and influencing architecture and execution beyond one's immediate workstream.
Strong track record delivering complex projects involving multiple stakeholders, dependencies, or phases.
Strong backend engineering fundamentals and proficiency in a language such as Ruby on Rails, Python, JavaScript, or Go.
SQL and relational database experience; PostgreSQL is preferred.
Hands-on experience with the AI application lifecycle, including experiment setup, dataset curation, offline evaluation, error analysis, deployment, and online evaluation.
Ability to operate amid ambiguity, frame problems, propose solutions, and align others.
Strong cross-functional communication skills, including explaining technical ideas to technical and non-technical audiences and translating complex concepts into business impact.
Ownership mindset and bias for action; raises the team's bar through mentoring, code review, and personal work.
Curiosity, continuous learning, and an agent-forward approach to software development.

Nice To Haves

Experience with pgvector or other vector indexes and query performance at scale.
Experience ingesting content from third-party systems such as document stores, knowledge bases, or CRM knowledge objects, including sync state, deduplication, and source conflicts.
Experience processing conversation or call transcripts, including PII and consent constraints.
Experience designing evaluation suites that measure qualities such as thoroughness and faithfulness, in addition to correctness.
Experience with voice or real-time conversational agents.
Experience with enterprise SaaS, especially B2B marketing or sales technology, or with Salesforce/CRM integrations.
Startup experience and rapid shipping.

Responsibilities

Lead retrieval initiatives from design through launch and operation, using RAG and hybrid search techniques including lexical and vector search, reranking, and filtering.
Build and drive technical design for developer frameworks and platform components for production agents, including multi-agent orchestration and shared primitives for prompts, tools, tracing, and evaluation.
Diagnose answer-quality issues by distinguishing retrieval problems from generation problems, define measurable quality standards, and lead improvements.
Design content and data models for ingesting, segmenting, labeling, and retrieving documents, knowledge bases, and call transcripts, including freshness, deduplication, source precedence, and visibility controls.
Build evaluation, observability, and rollout practices such as offline evaluations, online monitoring, and regression detection to make changes to prompts, models, and retrieval safe.
Evaluate solution paths and determine when to trust, redirect, or override model and agent output.
Set and track performance and scaling benchmarks for retrieval and indexing, LLM provider usage, and agent orchestration.
Own the reliability and performance of team systems, participate in on-call, lead incidents in the area, and communicate with stakeholders during incidents.
Use coding agents for implementation, testing, refactoring, and documentation, and create reusable workflows, patterns, and tools that improve team effectiveness.
Partner with product, design, and engineering leadership to shape roadmaps, tradeoffs, and requirements and turn strategy into execution plans.
Mentor engineers, provide code review and technical guidance, and improve team standards, tooling, and processes.