AI engineer (Sales tech)

Build the context + orchestration layer for AI first revenue software

Location: Hybrid in San Francisco, New York City or Vancouver

Work type: 3 days in office, 2 days remote

Our client is solving the biggest challenge companies face, growing revenue. Revenue teams squander great engineering because they prioritize or execute poorly. The team is tackling this by using AI agents that are built on top of our collective revenue memory.

The product connects messy, real world data into context that AI agents can learn from and act on autonomously. The goal: replace metawork with action that compounds revenue per rep, less digging through data and dashboards, more time moving pipeline.

If you're entrepreneurial and want to help define what AI native software looks like how data becomes context, how agents plan and act, and how chat turns into outcomes, you'll like building here. We ship in tight loops, learn in public, and measure by customer impact. Backed by SYN Ventures, our client is working with high growth cyber security and medtech customers like BigID and Modern Health. The team is product and engineering veterans from Salesforce, 6sense, Databook, Endgame, and Pitchbook; plus our board and advisors bring deep expertise from MIT, Symantec, DoD, Asana, RSA, Cylance, JPMC.

If this work gets you excited and you want to leave a mark, let's talk.

What you'll do

• Build the collective memory: Unify data from many sources (public, CRM and far beyond) into a semistructured context graph that captures what leads to winning dealsmodeled at multiple levels.
• Orchestrate agentic systems: Design planner/executor patterns, tools, and policies (including MCPstyle interfaces) that turn context into content and then into actions. Define simple eval harnesses to measure quality.
• Deliver where users work: Expose capabilities through native surfaces (apps, chat, and integrations) in tight loops with product and GTM, reducing context switches and metawork.
• Prove outcomes: With product and customers, define success metrics (e.g. tasks autocompleted, adoption/retention, pipeline lift; keep latency in check) and wire observability so we can ship,  learn iterate quickly.
• Balance cost & reliability: Tune accuracy, latency, and cost for agent runs and retrieval; design fallbacks and safeguards that keep the system dependable under real-world load.

What you'll bring

• Owner/builder mindset with product taste,  you frame problems, choose the simplest path, and own outcomes.
• 4+ years building & owning backend/platform systems end-to-end, with 01 wins and measurable business impact.
• Curious by default; comfortable taking smart risks and turning fuzzy problems into shipped outcomes.
• You talk in terms of impact and trade-offs; decide with ~70% info; turn ambiguity into simple, testable systems.
• Experience stitching messy, multisource data into something a product can reason over; strong instincts for reliability, privacy, and multitenant boundaries.
• Able to hit the ground running with LLMs, Python and handling cloud infrastructure.
• Nice to have: exposure to agent orchestration/planning, retrieval/graph-shaped context, eval frameworks, and distributed systems at scale.

How we work

• Outcome first: We anchor on the seller's job; stay close to customers; success = adoption, pipeline quality, time to value.
• Ship small, learn fast: Start simple; instrument; iterate with sniff tests.
• High trust, high ownership: Own problems end-to-end and make product-level decisions with the team.

Tech stack

We build with: Python, FastAPI/GraphQL, PostgreSQL/DynamoDB, AWS, Kubernetes, Pulumi, Spark/Databricks, and event-driven architectures; React for product surfaces. Familiarity helps, but isn't required.

Logistics & offer

• Equity: meaningful ownership in a fast-growing company
• Benefits: health, dental, vision
• Location: Hybrid San Francisco, New York City, Vancouver 3 days in office, 2 remote
• Team: small, senior; big surface area and ownership