Tech Lead Manager, Context

Profound•San Francisco, CA

7d•$230,000 - $275,000•Hybrid

About The Position

Profound is the marketing platform for the age of AI search. The way brands reach people is being rewritten — AI models like ChatGPT, Perplexity, and Google AI Mode are now the answer layer between companies and their customers. We built the platform marketers use to understand, measure, and win in that world: the analytics, intelligence, and agent automation that turn AI search from a threat into a competitive advantage. We went from 0 to a $1B valuation in 18 months. Revenue grew 100x last year. Our customers include Ramp, Figma, Spotify, Nike, Apple, AWS, Reddit, and JPMC. We are backed by Sequoia, Kleiner Perkins, LSVP, and Khosla Ventures — and we are moving fast enough that the people joining now are building the playbook everyone who comes after them will run. We are hiring a Tech Lead Manager to build and lead our Context team in San Francisco. Context owns everything Profound knows about a customer and everything our agents reach for when they do work. Every integration a customer connects, every document and page we ingest, every entity we resolve, and every piece of durable memory an agent carries between sessions runs through this team. If our agents are the hands, Context is the memory and the senses. The quality of what we ship is bounded by the quality of the context we retrieve. This is a hybrid role. You'll spend roughly half your time on technical leadership and architecture, and half on people management, hiring, and execution. We expect you to write code, review PRs, and be in the details on critical decisions. We also expect you to grow engineers, run a tight planning process, and own outcomes for the team. The Context team owns the integration service (the shared substrate for connecting to third-party systems: OAuth flows and token lifecycle, API key management, MCP servers and clients, credential storage and rotation, webhook ingestion, and per-provider rate limiting and backoff); the integration catalog itself (CMS platforms, analytics and search consoles, CRMs, document stores, support desks, and commerce systems, each with its own auth model, pagination quirks, and failure modes); knowledge base storage and retrieval (ingestion and chunking, embeddings, hybrid keyword and vector search, reranking, freshness, and strict tenant isolation across a store handling billions of documents); data synchronization (incremental sync, backfills, change detection, deduplication, conflict resolution, and the observability that tells us a customer's data is current before they notice it isn't); and context management (the knowledge graph that resolves brands, products, competitors, topics, and pages into a coherent model of a customer's world, plus the long-term memory service that lets agents accumulate and reuse what they learn).

Requirements

7+ years of backend engineering experience, with at least 2 years managing or tech-leading a team
Strong Python. This is the language the team lives in
Rust is strongly preferred. Parts of this system are performance and cost sensitive enough to justify it, and we expect that surface to grow
Hands-on experience building search or retrieval systems in production: embeddings, vector stores, hybrid search, reranking, and chunking strategies, along with an honest view of where each one stops helping
Experience integrating with third-party APIs at scale, including OAuth, token refresh, rate limits, and the operational reality of connectors that break on someone else's schedule
Experience with data synchronization and pipeline correctness: incremental sync, idempotency, backfills, and reconciliation
Strong Postgres skills; you can reason about indexes, query plans, multi-tenant data models, and migrations without reaching for a tutorial
Experience hiring senior ICs and giving direct, useful feedback
Comfortable being an IC contributor on the team you manage; we don't want a manager who has stopped writing code

Nice To Haves

Experience with knowledge graphs or entity resolution at scale
Experience building memory systems or context management for LLM agents, including context budgeting and retrieval tuning against real workloads
Familiarity with MCP as both a server and a client implementer
Experience with high-volume event ingestion and streaming systems
Background at a B2B SaaS company that sold to enterprise, where customer data handling passed a real security review (SOC 2, ISO 27001, CMEK, private networking)
A track record of building evaluation harnesses for non-deterministic systems

Responsibilities

Lead the Context roadmap, partnering with the CTO, Head of Engineering, and product leadership on priorities
Hire and develop a team of 4 to 8 engineers, including running interview loops, calibrating performance, and setting expectations clearly
Drive technical decisions on hard problems: retrieval quality and how we measure it, graph and memory schemas that survive contact with real customer data, sync architectures that stay correct under partial failure, and an integration framework that makes the tenth connector cheaper to build than the first
Own retrieval quality as a measurable product surface, with evals that catch regressions before customers do
Ship code yourself on the highest-leverage parts of the system
Be the engineering point of contact when a customer's integration breaks, their data goes stale, or an agent retrieves the wrong thing
Represent Context in conversations with product, sales engineering, and the executive team when data access, freshness, or security commitments are on the line
Partner closely with the teams building our agent products, who are your primary internal customers