Databricks resident solutions architect vs FDE: what the postings show and a candidate’s report

By FDE Interview Prep, a SourceStrongAI company

In this post11 sections

1. What happened to the resident solutions architect role
2. What Databricks’ FDE postings describe
3. The delivery solutions architect: the post-sale technical lead
4. The two jobs side by side
5. What Databricks publishes about its interviews (and what it does not)
6. Repositioning an architect’s resume for a Databricks FDE loop
7. If your process changes midway
8. Close the builder gap first
9. Questions people ask
10. Keep reading
11. More from the blog

Databricks now posts forward deployed engineer jobs, and its Professional Services page lists FDEs where resident architects used to be. ^{[1, 2, 3, 4, 5]}Source 1Sr. Staff Forward Deployed Engineer, SingaporePublisherDatabricks (Greenhouse)Source typecompany job postingSource 2Sr. Manager, AI Forward Deployed Engineering (AI FDE), United StatesPublisherDatabricks (Greenhouse)Source typecompany job postingSource 3Databricks Forward Deployed Engineering | DatabricksPublisherDatabricksSource typecompany websiteSource 4Databricks Professional Services (HTTP 301 to /forward-deployed-engineering)PublisherDatabricksSource typecompany websiteSource 5Databricks Professional Services | Databricks (Internet Archive)PublisherDatabricks (archived by the Internet Archive)Source typearchived company page One candidate for a resident solutions architect (RSA) role reported, in July 2026, that a recruiter told them mid-loop the role was becoming an FDE role. ^[6]Source 6Databricks Interview ExperiencePublisherBlind (teamblind.com)Source typecandidate report on Blind If you are an architect wondering which interview you face, here is the short answer. In May 2026, Databricks’ Professional Services page listed “Resident solutions architects” as its hands-on delivery experts; as of September 2026, that address opens the Forward Deployed Engineering page, which describes FDEs in the same words. ^{[3, 4, 5]}Source 3Databricks Forward Deployed Engineering | DatabricksPublisherDatabricksSource typecompany websiteSource 4Databricks Professional Services (HTTP 301 to /forward-deployed-engineering)PublisherDatabricksSource typecompany websiteSource 5Databricks Professional Services | Databricks (Internet Archive)PublisherDatabricks (archived by the Internet Archive)Source typearchived company page Databricks says the FDE organization brings Professional Services under one roof, so prepare for the builder’s interview. ^[7]Source 7Forward Deployed Engineering: Delivering Business Outcomes with AI (Jason Martin)PublisherDatabricks BlogSource typecompany blog The Delivery Solutions Architect is a different, post-sale planning role. For the comparison across other employers, read the FDE vs solutions architect guide.

What happened to the resident solutions architect role

Databricks has not announced, in anything we read, that it converted RSA jobs into FDE jobs. And we found no Databricks posting for an RSA, so we cannot quote its duties from a job description. What we can show is what Databricks’ own website said, before and after.

Before. An archived copy of the Professional Services page from May 2026 lists “Resident solutions architects” and data scientists as the experts behind its projects. ^[5]Source 5Databricks Professional Services | Databricks (Internet Archive)PublisherDatabricks (archived by the Internet Archive)Source typearchived company page It describes them as highly experienced technical resources with strong leadership and consulting skills, Databricks and Spark experts with a strong big data background and hands-on implementation skills. ^[5]Source 5Databricks Professional Services | Databricks (Internet Archive)PublisherDatabricks (archived by the Internet Archive)Source typearchived company page

After. As of September 2026, the Professional Services address redirects to Databricks’ Forward Deployed Engineering page. ^{[3, 4]}Source 3Databricks Forward Deployed Engineering | DatabricksPublisherDatabricksSource typecompany websiteSource 4Databricks Professional Services (HTTP 301 to /forward-deployed-engineering)PublisherDatabricksSource typecompany website That page lists “Forward Deployed Engineers (FDEs) and AI FDEs” with the same description: consulting skills, Spark expertise, hands-on implementation. ^{[3, 4]}Source 3Databricks Forward Deployed Engineering | DatabricksPublisherDatabricksSource typecompany websiteSource 4Databricks Professional Services (HTTP 301 to /forward-deployed-engineering)PublisherDatabricksSource typecompany website

In between. In June 2026, Databricks announced a Forward Deployed Engineering organization that brings its Professional Services organization “together under one roof”, and said FDE is not a new practice at the company. ^[7]Source 7Forward Deployed Engineering: Delivering Business Outcomes with AI (Jason Martin)PublisherDatabricks BlogSource typecompany blog

In our reading, the hands-on delivery role Databricks sold to customers has a new name and a new home. That is an inference from the pages, not a statement Databricks made about anyone’s job title.

The candidate report fits that picture. One Blind poster reported, in July 2026, that midway through a Resident Solutions Architect loop the recruiter said all RSA roles were being converted into FDE roles and asked for more rounds; the poster was then rejected without feedback. ^[6]Source 6Databricks Interview ExperiencePublisherBlind (teamblind.com)Source typecandidate report on Blind That is one person relaying what a recruiter said, so treat it as a signal, not a fact. Our lesson on the evidence map explains why we keep the company’s words and a candidate’s words in separate piles.

What Databricks’ FDE postings describe

Databricks’ FDE postings describe FDEs as billable, customer-facing builders who “own the architecture, lead design decisions, and implement end-to-end systems spanning data engineering, AI, and application development.” ^[8]Source 8Sr. Forward Deployed Engineer (FDE) - Financial Services (New York City)PublisherDatabricks (careers site / Greenhouse)Source typecompany job posting Read the verbs: own, lead, implement. That is delivery, and the customer pays for it.

The AI track says the same thing in AI terms. Databricks describes its AI FDE team as delivering professional services engagements to help customers build and productionize AI applications, and asks AI FDEs to own production rollouts of GenAI applications. ^[9]Source 9AI Engineer – Forward Deployed Engineering (AI FDE)PublisherDatabricks (Greenhouse)Source typecompany job posting

These lines tell you what the organization is for:

• Where it comes from. Databricks’ Sr. Deployment Strategist, FDE postings call the FDE team a new strategic initiative sponsored “from the CEO and Founders down”, whose mission is to win C-suite-sponsored, high-value AI opportunities “by selling and delivering outcomes, not just a platform.” ^[10]Source 10Sr. Deployment Strategist, FDE - Financial ServicesPublisherDatabricks (careers site / Greenhouse)Source typecompany job posting
• How the work is priced. The launch post describes increasingly outcome-aligned pricing with milestone-based, fixed-fee options. ^[7]Source 7Forward Deployed Engineering: Delivering Business Outcomes with AI (Jason Martin)PublisherDatabricks BlogSource typecompany blog Our advice: fixed-fee, milestone work means an FDE is judged on hitting delivery milestones, so bring one story where you committed to a date and a scope and met it, or renegotiated it early.
• How far the ladder goes. As of September 2026, Databricks’ postings include Forward Deployed Engineer, Sr. Forward Deployed Engineer, Staff and Sr. Staff titles, plus Manager and Senior Manager roles. ^{[1, 2]}Source 1Sr. Staff Forward Deployed Engineer, SingaporePublisherDatabricks (Greenhouse)Source typecompany job postingSource 2Sr. Manager, AI Forward Deployed Engineering (AI FDE), United StatesPublisherDatabricks (Greenhouse)Source typecompany job posting Pay is out of scope here; the FDE salary guide covers it.

The technical ask

The US Sr. FDE postings ask for 6+ years in data engineering, data platforms and analytics, or software engineering; deep Apache Spark experience, including knowledge of Spark runtime internals; coding in Python, Scala or JavaScript and TypeScript; and working knowledge of two or more clouds. ^[8]Source 8Sr. Forward Deployed Engineer (FDE) - Financial Services (New York City)PublisherDatabricks (careers site / Greenhouse)Source typecompany job posting If your recent work has been architecture, take “Spark runtime internals” most seriously. It means you can explain why a job is slow, not only which service to use.

The delivery solutions architect: the post-sale technical lead

The architect role Databricks does post after the sale is the Delivery Solutions Architect (DSA). Its posting describes a hybrid technical and commercial role. The holder becomes the post-sale technical lead across Databricks products, reports to a DSA Manager in the Field Engineering organization, works with Solutions Architects on customers’ use-case demand plans, and leads the technical account strategy after the technical win. ^[11]Source 11Delivery Solutions ArchitectPublisherDatabricks (Greenhouse)Source typecompany job posting

It is not a slide-only job. As of September 2026, the posting asks the DSA to plan how use cases move to production, to be the first contact for go-live issues, to coordinate with Professional Services on the delivery of its engagements, and to work with Product and Engineering on new features and private previews. ^[11]Source 11Delivery Solutions ArchitectPublisherDatabricks (Greenhouse)Source typecompany job posting It also asks for programming experience in Python, Spark and Databricks. ^[11]Source 11Delivery Solutions ArchitectPublisherDatabricks (Greenhouse)Source typecompany job posting

So both roles are technical. The difference is who delivers. The DSA leads the account’s plan and coordinates Professional Services on delivery; the FDE is the billable engineer who owns the architecture and the build. ^{[8, 11]}Source 8Sr. Forward Deployed Engineer (FDE) - Financial Services (New York City)PublisherDatabricks (careers site / Greenhouse)Source typecompany job postingSource 11Delivery Solutions ArchitectPublisherDatabricks (Greenhouse)Source typecompany job posting

The two jobs side by side

This compares what the postings and pages say. It is not a claim about how either team works day to day.

In our reading, the “Delivery” row is the real split, and the two rows under it follow from it. The DSA makes sure the work happens; the FDE does the work. Bring evidence that you built a system yourself and took it into production, and that you handled the moment the platform could not do what the customer needed. Our lesson on FDE, SWE, solutions architect and sales engineer roles walks through the same split at other employers.

What Databricks publishes about its interviews (and what it does not)

We found no published Databricks FDE interview loop. Here is what Databricks does publish, and what each one covers.

• The company-wide process. Databricks’ careers page says its process generally includes a recruiter call, a pre-onsite screen, an onsite loop and, for some roles, a presentation. ^[12]Source 12Our hiring process / Interview prepPublisherDatabricks CareersSource typecompany hiring page That covers every role, not the FDE loop.
• The go-to-market process. Databricks’ Go to Market careers page lists a hiring manager screen, a technical interview, a virtual interview loop, a virtual presentation and references, and says the process varies by role. ^[13]Source 13Go to Market at DatabricksPublisherDatabricks CareersSource typecompany job posting The page does not say which roles it covers.
• The nearest role-specific loop, for a different role. A Delivery Solutions Architect posting in Tokyo, which is not an FDE role, lists a recruiter screen, a hiring manager screen, “Design & Architecture”, “Vibe Coding”, “Build, Demo & Delivery” and a reference check. ^[14]Source 14Delivery Solutions Architect (Tokyo, Japan)PublisherDatabricks (careers site / Greenhouse)Source typecompany job posting

Even that architect loop lists coding and a build. So do not assume an architect background exempts you from writing code live. Our advice: prepare for at least as much in an FDE loop, since the FDE posting puts the build in your hands.

One practical rule applies to every Databricks role: its interview best practices tell candidates not to use their current work computer or current employer’s materials when creating a candidate assignment or presentation. ^[12]Source 12Our hiring process / Interview prepPublisherDatabricks CareersSource typecompany hiring page If you work at a partner or customer today, build your demo from scratch on your own machine.

What to rehearse for the builder half

Since the postings ask for Spark runtime internals, prepare to explain these without notes. The Spark RDD programming guide and the Spark SQL performance tuning page are the primary sources.

A slow join, narrated

Here is how that list sounds when you use it. The scenario is invented.

A nightly job that joins orders to customers has started missing its deadline. In the Spark UI, the join stage shows nearly every task finishing in seconds and a single task running for most of the job’s wall time.

1. Open the slow stage and compare task durations. A long tail on one task, with the rest fast, points at skew, not at too little compute.
2. Read the plan. joined.explain() shows a SortMergeJoin with an Exchange hashpartitioning on customer_id under each side. So every row with the same key goes to the same task.
3. Count rows per key.

from pyspark.sql import functions as F
(orders.groupBy("customer_id").count()
    .orderBy(F.desc("count")).show(5))

One customer_id holds most of the rows: a test account, or a default value filled in upstream.

4. Fix it. If customers is small, broadcast it: orders.join(F.broadcast(customers), "customer_id"). The plan now shows BroadcastHashJoin and a BroadcastExchange, and the large side is not shuffled. If both sides are large, check that adaptive query execution’s skew-join handling (spark.sql.adaptive.skewJoin.enabled) is on, which splits oversized partitions in a sort-merge join. If the hot key is junk, filter it out and ask who writes it.

The sentence to say out loud: “One task was doing most of the work, the plan showed a sort-merge join on customer_id, and one key held most of the rows, so I broadcast the small side and the stage evened out.”

Then practice a design prompt in the same shape as the work. Designing a unified retail data platform makes you merge point-of-sale, e-commerce, loyalty and supplier feeds: the kind of lakehouse build an FDE owns end to end. If you are preparing for Databricks’ AI track instead, the Databricks AI FDE coding round post covers what one candidate was told about it.

Repositioning an architect’s resume for a Databricks FDE loop

For the full architect-to-FDE rewrite, see closing the production-code gap. For Databricks specifically, three kinds of bullet carry the most weight, because they match the postings: Spark and Delta work you did yourself, a billable delivery with milestones, and a platform-gap story. ^{[7, 8]}Source 7Forward Deployed Engineering: Delivering Business Outcomes with AI (Jason Martin)PublisherDatabricks BlogSource typecompany blogSource 8Sr. Forward Deployed Engineer (FDE) - Financial Services (New York City)PublisherDatabricks (careers site / Greenhouse)Source typecompany job posting

• Before: “Advised a retailer on lakehouse migration.”
  After: “Built the ingestion jobs that moved a retailer’s nightly loads onto the lakehouse, and cut the run from [X] hours to [Y] minutes, so reports landed before the morning meeting.”
• Before: “Led architecture reviews for strategic accounts.”
  After: “Found and fixed a skewed join stalling a customer’s pipeline, then added a skew check that failed the deploy when one key held more than [N]% of rows.”
• Before: “Scoped a fixed-price engagement.”
  After: “Delivered a fixed-scope engagement across [N] milestones, and renegotiated one early when the source system slipped.”

Keep the architect strengths: scoping, trade-off calls and executive trust matter in an FDE loop too. Just make each one sit on top of something you shipped. The FDE resume guide goes through the whole document.

The stories to rehearse

Rehearse working inside a customer’s environment first, because it asks for exactly the hands-on evidence an architect’s resume hides. Then prepare custom build or product change: Databricks says its FDEs go to R&D when the platform falls short, so have one story where you made that call. ^[7]Source 7Forward Deployed Engineering: Delivering Business Outcomes with AI (Jason Martin)PublisherDatabricks BlogSource typecompany blog

Mistakes that cost architects the builder rounds

• Drawing boxes instead of writing code. In a coding or build round, a diagram is not an answer. Get something running first, then explain it.
• Talking in “we”. Say “I” for what you built and decided. Interviewers need to know which part was yours.
• Staying at the product level. “We used Delta and Unity Catalog” is a list. “I used liquid clustering on customer_id and event_date instead of date partitions, because the table was under a terabyte and queries filtered on both” is a decision, and it matches Databricks’ own partitioning guidance, which says not to partition tables under a terabyte and recommends liquid clustering for managed tables.
• Skipping the failure. Every production story has a bad night. Tell it, and say what you changed afterward.

If your process changes midway

The Blind report above is a useful warning even though it is one account: the process changed partway through, and no feedback came. ^[6]Source 6Databricks Interview ExperiencePublisherBlind (teamblind.com)Source typecandidate report on Blind If a recruiter tells you your role is becoming a different one, ask before you prepare another minute.

Write down the answers and the date. If the new role is an FDE role, re-read its posting against the table above and move your preparation toward the builder rows: production code, Spark internals and a story about extending the platform. If the answers are vague, that is also information. Our post on reading FDE interview reports on Reddit and Blind shows how to weigh one person’s account without over-reading it.

Close the builder gap first

You already know the architecture half. Start with the free lesson what FDE coding rounds test, then work the unified retail data platform against its model answer. Pro opens every module on the rounds, with a 7-day free trial on the pricing page.

Questions people ask

Is Databricks converting resident solutions architect roles into FDE roles, as one Blind poster reported?

One Blind poster reported, in July 2026, that a recruiter told them so. Databricks has not said so in anything we read, but its own pages point the same way. In May 2026 its Professional Services page listed resident solutions architects as its hands-on experts. As of September 2026 that address opens the Forward Deployed Engineering page, which lists FDEs in the same words, and Databricks says the FDE organization brings Professional Services under one roof.^{[3, 4, 5, 6, 7]}Source 3Databricks Forward Deployed Engineering | DatabricksPublisherDatabricksSource typecompany websiteSource 4Databricks Professional Services (HTTP 301 to /forward-deployed-engineering)PublisherDatabricksSource typecompany websiteSource 5Databricks Professional Services | Databricks (Internet Archive)PublisherDatabricks (archived by the Internet Archive)Source typearchived company pageSource 6Databricks Interview ExperiencePublisherBlind (teamblind.com)Source typecandidate report on BlindSource 7Forward Deployed Engineering: Delivering Business Outcomes with AI (Jason Martin)PublisherDatabricks BlogSource typecompany blog

What did a Databricks resident solutions architect do?

We found no Databricks posting for the role. An archived copy of Databricks’ Professional Services page from May 2026 describes resident solutions architects as highly experienced technical resources with strong leadership and consulting skills, Databricks and Spark experts with hands-on implementation skills. As of September 2026 the same page lists FDEs with that description.^{[3, 4, 5]}Source 3Databricks Forward Deployed Engineering | DatabricksPublisherDatabricksSource typecompany websiteSource 4Databricks Professional Services (HTTP 301 to /forward-deployed-engineering)PublisherDatabricksSource typecompany websiteSource 5Databricks Professional Services | Databricks (Internet Archive)PublisherDatabricks (archived by the Internet Archive)Source typearchived company page

What does a Databricks FDE do?

Databricks’ postings describe FDEs as billable, customer-facing builders who own the architecture and implement end-to-end systems across data engineering, AI and application development. Its AI FDE team delivers professional services engagements and owns production rollouts of GenAI applications.^{[8, 9]}Source 8Sr. Forward Deployed Engineer (FDE) - Financial Services (New York City)PublisherDatabricks (careers site / Greenhouse)Source typecompany job postingSource 9AI Engineer – Forward Deployed Engineering (AI FDE)PublisherDatabricks (Greenhouse)Source typecompany job posting

Do Databricks FDE roles need Spark?

The US Sr. FDE postings ask for deep Apache Spark experience, including knowledge of Spark runtime internals, alongside coding in Python, Scala or JavaScript and TypeScript.^[8]Source 8Sr. Forward Deployed Engineer (FDE) - Financial Services (New York City)PublisherDatabricks (careers site / Greenhouse)Source typecompany job posting

Keep reading

Guides

• Forward deployed engineer vs solutions architect
• The forward deployed engineer resume
• Forward deployed engineer salary: what employers post and file

Company guides

• Databricks

Lessons

• FDE, software engineer, solutions architect, sales engineer: telling them apart
• The evidence map: what employers publish, what candidates report, how to read both
• What FDE coding rounds test, by the evidence

Questions

• Tell me about a time you worked inside a customer’s environment rather than your own. What was hardest?
• Design a data platform that unifies a retailer’s point-of-sale, e-commerce, loyalty and supplier feeds into one customer and product view.
• How do you decide between building a custom solution for one customer and pushing for a product change?

More from the blog

1. Company loops

   Google Cloud customer engineer vs forward deployed engineer: same company, different jobs

   Google Cloud customer engineer vs forward deployed engineer: what each posting says the job is, how pay is structured, and what candidates report.

   10 min readPublished September 2026

2. Company loops

   The OpenAI Deployment Company: what it is and what it means for FDE candidates

   The OpenAI Deployment Company (DeployCo) embeds FDEs with customers. Where it hires, what its postings ask, how engagements start and how to prepare.

   10 min readPublished September 2026

3. Company loops

   Databricks AI FDE coding round: preparing for applied data science and machine learning

   What Databricks publishes, what one candidate was told about the AI FDE coding round, and a prep plan: a baseline model, leakage checks and trade-offs.

   10 min readPublished September 2026

More in Company loops