Search Engineer
Full-time
On-site
San Francisco, CA, US
$235K - $260K
Role highlights
Requirements
Pipeline
About this role
What to ask
Interview process
Ideal candidates
Rejected feedback
About Firecrawl
Announcements
Similar roles
Apply to recruit
Out of credits
0 credits left
Ankit Bahuguna
Ankit Bahuguna
Head of Search
Highly engaged
Top 1% of hiring managers
Time to first action
9 hours
Reply time
37 hours
Active days
9 days
Feedback quality
Clear
Latest announcement
· Sep 25, 2026
Hi Team,Quick update on Search Engineer. We rewrote the role page, and the biggest change is a correction, so please read this one closely.The bar: ranking and relevance as the candidate's primary, owned muscle. They've shipped ranking or re-ranking changes (BM25, semantic/vector retrieval, re-ranking, RAG) that moved a production metric like relevance, recall, or conversion, and they can name the number.What changed: the old sourcing advice called this role systems-first and said not to over-index on ranking depth. That was wrong, and it's why most declines on this role have been strong backend engineers. Backend, indexing, and latency work are a plus on top of owned ranking, never a substitute for it.Where to look: ranking and relevance teams at web search engines, e-commerce and marketplace search, enterprise search, and retrieval-quality teams at AI companies. The team matters more than the company name.What doesn't clear:Strong backend or distributed-systems engineers with no owned ranking workSearch infra or middleware (OpenSearch, Lucene, ingestion pipelines, feature stores)Retrieval serving where only latency or cost movedVector search wired into an app or side projectAlso new:Location: on-site five days a week in SF or Toronto (Toronto is new). Candidates can relocate before starting. No fully remote candidates.Work authorization: must be authorized to work in the US or Canada. No US visa sponsorship right now; Canada is case by case.Base: $235K-$260K USD (SF) or $217K-$233K CAD (Toronto), plus competitive equity. The band is firm, so ignore the old note about room above $260K.Feedback: Ryan Jimenez, our Deputy Chief of Staff, now reviews every profile. Same clear yes/no with reasons.Paid work trial: ~40 hours.For your outreach: we just announced a $75M Series B and we're at 180k+ GitHub stars (top 50 repos of all time). Updated selling points and outreach guidelines are on the role page.If you have profiles in progress, run them against the new bar before submitting. Looking forward to what you send.Mark
Role highlights
Hiring goal
1 candidate
Recruiter activity
Recruiter activity on this role is medium
No final round interviews
There currently aren't any candidates in final round interviews or offer stage
Why candidates should join
Rocketship growth, real technical depth. 8 figures ARR in year one, more than doubled in year two. 180k+ GitHub stars, top 50 repos of all time. $75M Series B announced September 2026, led by Smash Capital with Altos Ventures, Nexus Venture Partners, Y Combinator, and more.
Own ranking for a fast-growing search product. You'll own ranking quality and relevance for Firecrawl Search, one of our fastest-growing products, plus the crawling, indexing, and retrieval systems that serve it. Billions of docs, real latency and cost constraints.
Small team, outsized impact. A ~5-person search team. You'll work directly with the Head of Search and shape the roadmap. No committees, no waiting.
Build what superintelligence will use. The search systems you build serve the next generation of AI agents and models. Core product, not a side feature.
Comp and benefits. $235K-$260K USD (SF) or $217K-$233K CAD (Toronto) base, plus competitive equity. On-site five days a week at SF HQ or our new Toronto Hub, with 100% employer-paid health coverage for you, 12 weeks paid parental leave, $1K/yr L&D, and a 3-month paid sabbatical after 4 years.
Explosive Growth
Explosive growth, real revenue. 8 figures ARR in year one, more than doubled in year two. $75M Series B announced September 2026, led by Smash Capital with Altos Ventures, Nexus Venture Partners, Y Combinator, and more.
Developers Already Love It.
Developers already love it. 180k+ GitHub stars (top 50 repos of all time) and category-defining AI companies building on it daily. Your demos start with credibility.
The Data Layer for Superintelligence
Every AI lab, every agent, and every model needs clean web data, and that demand is exploding right now. Firecrawl is building the infrastructure that superintelligence will rely on to gather data from the web. You'd be joining at the exact moment the category is forming, working on a problem that gets more central to AI every month, not less.
Suggested outreach
Requirements
By tier
By category
Calibrate a candidate in Parabot. to see how well they fit the requirements.
Dealbreaker
Work experience
Can point to a specific ranking or re-ranking change they owned and the relevance, recall, or conversion metric it moved
Updated
Hard skills
Built or owned a production search engine's indexing and serving layers (index build, sharding, serving, freshness, query latency and QPS) at real scale as their main job for 3+ years, and can name the scale
Updated
Baseline
Seniority
3+ years of experience owning ranking and relevance in production as the primary discipline
Work experience
Owned search services end to end (design, ship, monitor, iterate in production), including latency and cost at high query volumes
Updated
Hard skills
Hands-on relevance depth on top of the engine they own: BM25, hybrid/semantic/vector retrieval, re-ranking, or RAG evaluation.
Relevance work without owning the engine is not enough
Updated
Backend systems programming (Go, Rust, Python, or equivalent)
Updated
Miscellaneous
On-site five days a week in SF or Toronto, now or by relocating before start
Updated
Nice-to-have
Work experience
Worked at a company where search is the product: web search, AI search or answer engines, or large e-commerce or marketplace search
Early-stage or high-growth startup experience
Updated
Traits to avoid
Strong backend, distributed-systems, or platform engineer with no personally owned ranking or relevance work, however strong the overall engineering
Updated
Relevance or ranking ML built on top of an existing search stack (embeddings, two-tower, cross-encoders, learning-to-rank) without owning the index or serving layer
Search-adjacent only: tuned or operated someone else's OpenSearch/Elasticsearch cluster, ingestion pipelines, feature stores, or "wired a vector DB into an app
Updated
Only moved latency, throughput, or cost, with no relevance metric they can name
Updated
Pipeline
Inbox
1candidate
Intro Chat
2candidates
Technical Chat
0candidates
Founders Chat
0candidates
Paid Work Trial
0candidates
Offer
0candidates
Hired
0candidates
Recent Activity
Changed status of Candidate #4 from Submitted to First round
3 days ago
Removed requirement (Nice-to-have): Applied ML or relevance modeling (ranking models, embeddings, learning-to-rank)
5 days ago
Requirement description updated: Search-adjacent only: tuned or operated someone else's OpenSearch/Elasticsearch cluster, ingestion pipelines, feature stores, or "wired a vector DB into an app
5 days ago
Added requirement (Required): Relevance or ranking ML built on top of an existing search stack (embeddings, two-tower, cross-encoders, learning-to-rank) without owning the index or serving layer
5 days ago
Added requirement (Nice-to-have): Worked at a company where search is the product: web search, AI search or answer engines, or large e-commerce or marketplace search
5 days ago
Requirement type: Must-have → Required, description updated: Hands-on relevance depth on top of the engine they own: BM25, hybrid/semantic/vector retrieval, re-ranking, or RAG evaluation. Relevance work without owning the engine is not enough
5 days ago
Requirement type: Nice-to-have → Must-have, description updated: Built or owned a production search engine's indexing and serving layers (index build, sharding, serving, freshness, query latency and QPS) at real scale as their main job for 3+ years, and can name the scale
5 days ago
Added requirement (Required): On-site five days a week in SF or Toronto, now or by relocating before start
5 days ago
About this role
Search Engineer at Firecrawl
You'll build the systems that let anyone turn the open web into a search index. Firecrawl's search product is one of our fastest-growing surfaces, and we need engineers who can make crawling, ranking, and retrieval fast, reliable, and cheap at scale. You'll own real infrastructure from day one, not tickets in a backlog.
About Firecrawl
Firecrawl is the easiest way to turn the web into data AI agents can use. One API call converts any URL into clean, LLM-ready markdown or structured data. It's the boring-hard problem everyone building with LLMs eventually hits, solved.
In September 2026 we raised a $75M Series B led by Smash Capital, and we're spending it building the largest repository of knowledge in the world. We hit 8 figures in ARR in year one and more than doubled it in year two. We have 183k+ GitHub stars, putting us in the top 50 repositories of all time, and developers, agents, and category-defining AI companies build on us every day. Growth like this is rare, and we're just getting started.
We're a small team punching far above our weight, working out of SF HQ and our new Toronto Hub. Everyone here owns a real piece of the product and company, end to end, and runs it themselves. No hiding behind process or headcount.
This is a place for people who want to work at the frontier: an AI company building the infrastructure other AI companies run on, not one bolting AI onto an existing product. We move fast, go deep, and are building the tools superintelligence will rely on to gather data from the web. That library is called Alexandria, and it starts now.
What You'll Do
Own ranking quality and relevance for LLM-driven retrieval. This is the core of the role
Design and build the crawling, indexing, and retrieval systems behind Firecrawl Search
Push down latency and cost per query while search volume grows
Own services end to end: design, ship, monitor, and iterate in production
Work directly with the Head of Search and the rest of the search team on the roadmap
What We're Looking For
3+ years building or owning a production search engine's indexing and serving layers at real scale, with hands-on relevance work on top (BM25, semantic and vector retrieval, re-ranking, RAG)
You've shipped ranking changes that moved a production metric (relevance, recall, conversion) and can point to the number
You care about latency, cost, and correctness in equal measure
You're comfortable owning ambiguous problems and turning them into shipped systems
You move fast and close the loop. You'd rather ship, measure, and iterate than perfect on paper
What We're NOT Looking For
Backend or distributed-systems engineers whose search work is infrastructure feeding someone else's ranking
Retrieval serving where latency moved but relevance didn't
Vector search wired into an app or side project, without owned relevance work
Someone who needs a fully-specced ticket to start
A Note On Pace
We operate at an absurd level of urgency because the window for what we're building won't stay open forever. If that excites you, keep reading. If it doesn't, no hard feelings, but this role probably isn't for you.
Role details
Salary
$235K - $260K
On-site work policy
San Francisco, CA (SF HQ) or Toronto, ON (Toronto Hub). On-site, five days a week.
Full-time position
Location
San Francisco, CA, US, Toronto, ON, Canada
Tech stack
Distributed Systems, Go, Rust, Python, Elasticsearch, Apache Lucene, Redis, Kafka, PostgreSQL, Kubernetes, Docker, AWS, GCP, BM25, Vector Search, Web Crawling Infrastructure, Indexing Pipelines, gRPC, Protobuf
Equity
Competitive equity — details shared during the process.
Visa sponsorship details
Open to visa transfers (e.g. OPT, H1B transfers)
Additional visa sponsorship details
Open to discussion for truly exceptional candidates, but not standard practice currently.
Hiring Count
Looking to hire 1 candidate for this role
Report to
https://www.linkedin.com/in/ankitbahuguna/
What to ask candidates
Can the candidate be on-site? If not, is the candidate willing to relocate?
What is their salary expectation?
How actively is this candidate exploring new opportunities?
Requirements
By tier
By category
Calibrate a candidate in Parabot. to see how well they fit the requirements.
Dealbreaker
Work experience
Can point to a specific ranking or re-ranking change they owned and the relevance, recall, or conversion metric it moved
Updated
Hard skills
Built or owned a production search engine's indexing and serving layers (index build, sharding, serving, freshness, query latency and QPS) at real scale as their main job for 3+ years, and can name the scale
Updated
Baseline
Seniority
3+ years of experience owning ranking and relevance in production as the primary discipline
Work experience
Owned search services end to end (design, ship, monitor, iterate in production), including latency and cost at high query volumes
Updated
Hard skills
Hands-on relevance depth on top of the engine they own: BM25, hybrid/semantic/vector retrieval, re-ranking, or RAG evaluation.
Relevance work without owning the engine is not enough
Updated
Backend systems programming (Go, Rust, Python, or equivalent)
Updated
Miscellaneous
On-site five days a week in SF or Toronto, now or by relocating before start
Updated
Nice-to-have
Work experience
Worked at a company where search is the product: web search, AI search or answer engines, or large e-commerce or marketplace search
Early-stage or high-growth startup experience
Updated
Traits to avoid
Strong backend, distributed-systems, or platform engineer with no personally owned ranking or relevance work, however strong the overall engineering
Updated
Relevance or ranking ML built on top of an existing search stack (embeddings, two-tower, cross-encoders, learning-to-rank) without owning the index or serving layer
Search-adjacent only: tuned or operated someone else's OpenSearch/Elasticsearch cluster, ingestion pipelines, feature stores, or "wired a vector DB into an app
Updated
Only moved latency, throughput, or cost, with no relevance metric they can name
