×
Feldera
Incremental view maintenance engine that recomputes only what changed
Last verified September 22, 2026 · Updated daily
What Feldera is building
Feldera is an Incremental View Maintenance (IVM) engine. You write standard SQL views over your tables (joins, aggregates, group-bys, even recursive queries), and instead of re-running the whole query when data changes, Feldera compiles the SQL into an incremental program whose work is proportional to the size of the change, not the size of the dataset. A view that used to refresh as a nightly batch job refreshes in seconds, continuously. The math underneath is DBSP, a framework for incremental computation over relational algebra that two of the co-founders (Mihai Budiu and Leonid Ryzhyk) co-created with Frank McSherry and Val Tannen. The DBSP paper won Best Paper at VLDB 2023. The pitch to a data platform team is that incremental correctness comes from the theory, not from hand-written stream-processing code that approximates the batch result. The engine is open source (github.com/feldera/feldera, Rust) with a cloud sandbox at try.feldera.com and an enterprise deployment that sits on top of an existing lakehouse or warehouse rather than replacing it. Production workloads named in the announcement: real-time fraud detection, logistics views over billions of dollars of operations, and fine-grained authorization for AI agents. The company's framing of the cost problem is blunt: most enterprise workloads are complex enough that teams have no choice but to rerun batch jobs 24/7, even when 99.9% of the data hasn't changed. Feldera claims 95%+ reductions in warehouse compute for those workloads, with average infrastructure cost reductions of 10x and some workloads over 100x.
Why this matters
Incremental view maintenance is a decades-old database problem that never got solved for general SQL; every practical system either restricted the query language (materialized views with narrow support) or asked engineers to hand-build streaming pipelines that drift from the batch truth. DBSP is the first approach with a general, proof-backed answer, and the VLDB 2023 Best Paper award is external validation of that, not marketing. The timing argument is about agents. Lalith Suresh puts it directly: AI agents are only as good as the data they can access, and if that data is stale because it's too expensive to recompute, companies burn tokens on agents reasoning over yesterday's state. Fine-grained authorization for agents (which principal can touch which row, right now) is itself an incremental-view problem. Inovia's stated reason for leading is the same thesis: as enterprises scale real-time AI, the cost and latency of traditional compute become unsustainable, and Feldera is the compute layer underneath. A logo list that already includes Okta (Auth0), Nubank, Procore and Zeta Global says the technology has left the paper.
Open roles at Feldera
4 positions we're tracking. Roles are re-checked daily and removed when filled.
Rust Systems Engineer (core DBSP engine, query compiler)
First seen today
Backend Infrastructure Engineer (connectors, Kafka/Iceberg/Delta integrations)
First seen today
Forward Deployed Engineer (customer pipeline architecture)
First seen today
Enterprise Account Executive (data platform buyers)
First seen today
Know when Feldera is hiring before anyone else
A role stays uncontested for about four days. Here's the window — and where we put you in it.
From $5.99/week, cancel any time.
Watching Feldera
0 applicantsRole spotted & verified
1You get the alert
1You've applied
~8Hits the job boards
250+Hiring outlook
The Series A announcement says the company is actively recruiting across departments, and the careers page describes a global, remote-first team that wants candidates to introduce themselves at workwithus@feldera.com. No individual job posts are enumerated, so treat the roles below as directional.
Working at Feldera
Feldera is a AI company based in San Francisco, USA. For a AI company this size, the reality is opportunity to shape your role based on the company stage.
Most Feldera jobs are based in San Francisco, USA.
How to actually get hired at Feldera
Why applying the normal way doesn't work
Feldera has an ATS but the real pipeline is referrals → sourced → recruiter picks → cold apps. You can still get through, but know where you stand.
Who to contact at Feldera
What to show them
Clone github.com/feldera/feldera, run a non-trivial SQL pipeline (a multi-way join plus a windowed aggregate, or a recursive query) through it, and write up one thing you found: a performance cliff, a correctness edge case, an operator you'd add, or a connector gap. A PR or a well-reasoned issue against the repo is the strongest possible artifact.
A cold email that works at Feldera
What Feldera screens for
Inovia leading a US-based database infrastructure Series A, with Battery joining and Ion Stoica already in from seed, is a bet on a research team that shipped: the engine is open source, the sandbox is public, and the customer list has Okta and Nubank on it. The stated plan (managed platform experience, deeper lakehouse integrations) means the next hires are platform engineers who can make a research-grade Rust engine boring to operate, plus the go-to-market people to sell it into data teams. The team is small enough that an emailed introduction with a real Feldera pipeline attached will be read by a co-founder.
Don't send a generic CV to Feldera. Mirror the job posting's language to get past automated screening.
Don't make these mistakes
Do not pitch 'streaming' or 'real-time data' as if it's new to them. This is a team of five former VMware Research scientists with 200+ papers between them, including the co-creators of DBSP and the founding developer of Open vSwitch. 'I'm passionate about real-time analytics' will get ignored. Also do not conflate Feldera with Flink or Materialize without having read why DBSP differs; Leonid and Mihai wrote the theory. Lead with something you actually ran in the open-source engine or the sandbox, and a specific incremental-correctness or performance observation from it.
Mistakes that kill Feldera applications
At people, a copy-paste CV is immediately obvious. It's an instant no.
Lead with their problem, not your ambition. Feldera is focused on Incremental view maintenance is a decades-old database problem that never got solved for general SQL; every practical system either restricted the query language (materialized views with narrow support) or asked engineers to hand-build streaming pipelines that drift from the batch truth — show you understand that.
Don't apply and wait. The median AI application gets no response ever. One follow-up at day five roughly doubles reply rates.
Applying to Feldera? Get the contact, not the form.
The Feldera interview process
4 stages · 14 days typical · take-home: yes · modelled from similar companies
We don't yet have verified candidate reports for Feldera. What follows is the typical process for a -person AI company — treat it as a model, not confirmed detail.
Interview stages
Recruiter Screen
Phone or video · 30 min
Basic qualification and logistics
Recruiter or HR
Hiring Manager Interview
Video call · 45 min
Role fit and experience deep-dive
Hiring manager
Technical/Functional Round
Video call · 60 min
Skills assessment and problem-solving
Team members
Final Round
In-person or video · 60 min
Culture fit and cross-functional alignment
Senior leadership
Feldera take-home assignment
Feldera includes a take-home exercise in their interview process. For AI roles, this typically involves a practical problem that takes 2-4 hours. Focus on clean, working code over premature optimization. They're evaluating how you think and communicate, not just the solution.
Feldera interview timeline
At days, Feldera's process is about average than typical for AI (14 days at this size).
Interviewed at Feldera?
Tell us how it went — stages, questions, timeline. Takes 90 seconds and it's how this page stays accurate for the next person.
Submit your Feldera interview experience →Feldera jobs, frequently asked questions
How many jobs does Feldera have open?
Feldera currently has 4 open roles, last verified September 2026.
Does Feldera hire remotely?
Currently, Feldera only has in-office roles in San Francisco, USA.
What roles is Feldera hiring for?
Feldera is hiring across Engineering. The most recent opening is Rust Systems Engineer (core DBSP engine, query compiler).
How do I apply for a job at Feldera?
Apply directly through the links above, or read our guide on how to actually get hired at Feldera.
Does Feldera respond to cold emails?
We're still collecting cold email data for Feldera.
Who is the hiring manager at Feldera?
At this size, hiring is usually run by the hiring manager.
How competitive is it to get hired at Feldera?
Expect 100-250 applicants in the first two weeks for AI roles at this size. Apply within 72 hours for best odds.
How many rounds is the Feldera interview?
4 stages: Recruiter Screen, Hiring Manager Interview, Technical/Functional Round, Final Round.
Is the Feldera interview hard?
It concentrates on technical depth and system design rather than abstract puzzles. The stage candidates find hardest is Technical Interview.
Does Feldera give a take-home task?
Yes, Feldera includes a take-home assignment.
How long does Feldera take to get back to you?
Around 14 days across the full process.
What should I prepare for the Feldera interview?
Study technical depth and system design. At this size (), they care about self-sufficiency over textbook knowledge.
Where is Feldera based?
Feldera is headquartered in San Francisco, USA.
Get Feldera roles before they're posted
A role stays uncontested for about four days. Here's the window — and where we put you in it.
From $5.99/week, cancel any time.
Watching Feldera
0 applicantsRole spotted & verified
1You get the alert
1You've applied
~8Hits the job boards
250+Related
More AI companies hiring
Other startups we track in the same space.