Interview guides · Deepgram

Deepgram Software Engineer Interview Guide

Deepgram Software Engineer interview guide: reported rounds, questions and how candidates describe the process.

The 60-second brief

As a Software Engineer at Deepgram, you will build and scale the backbone of next-generation voice AI and automated order-taking platforms. This role sits at the intersection of high-performance backend systems, real-time audio processing, and cutting-edge machine learning pipelines. Whether you are developing robust infrastructure for restaurants, optimizing ASR engines in noisy environments, or integrating complex third-party point-of-sale systems, your code directly impacts how millions of users interact with AI voice agents. The work requires tackling complex technical challenges that span distributed systems, cloud infrastructure, and low-latency API design. You will collaborate closely with core research teams, product owners, and machine learning engineers to push the boundaries of LLMs and voice technology.

This guide is scoped to a Software Engineer candidate at Deepgram.

Focus areas

Data Structures Fundamentals
Technical Screen (experience-based)
Take-Home Assignment Development

The Deepgram interview process

Deepgram candidates report 4 rounds over 3-5 weeks. The stages below are what candidates describe, not a published process.

Recruiter Screen

Initial discussion to align on timelines and background.

What to demonstrate

Initial discussion to align on timelines and background
Depth in Data Structures Fundamentals

How to prepare

Be able to walk your CV end to end in two minutes, and say why this company specifically.
Have your salary expectations, notice period and location constraints ready, and ask for the rest of the loop in writing.

Technical Screen

Discussion with a senior engineer about past projects and technical depth, without live coding.

What to demonstrate

Discussion with a senior engineer about past projects and technical depth, without live coding
Depth in Data Structures Fundamentals

How to prepare

Answer aloud and timed: What experience do you have with cloud-based infrastructure and deployment technologies like AWS?
Answer aloud and timed: How would you design an API that integrates with multiple third-party point-of-sale systems?

Take-Home Assignment

Build a basic application or API based on provided requirements.

What to demonstrate

Build a basic application or API based on provided requirements
Depth in Data Structures Fundamentals

How to prepare

Answer aloud and timed: What are the primary bottlenecks when scaling backend systems in production environments?
Answer aloud and timed: How would you design a real-time, menu-aware ordering system that handles ambiguous customer requests?

Virtual Onsite

Concise session including a live extension of the take-home assignment and a behavioral interview.

What to demonstrate

Concise session including a live extension of the take-home assignment and a behavioral interview
Depth in Data Structures Fundamentals

How to prepare

Answer aloud and timed: How do you approach zero-to-one backend system design for a new product initiative?
Answer aloud and timed: How would you architect a backend to reliably ingest and process audio streams in noisy environments?

Where candidates lose points

Clarify requirements early

Because some engineering exercises involve open-ended instructions, always confirm your understanding of the core requirements with your interviewers before writing code.

Focus on trade-offs

When discussing system design or past projects, never present a single solution as perfect; always articulate the trade-offs regarding latency, scale, and maintenance.

Highlight production experience

Ground your answers in real-world production scenarios, emphasizing how you handled monitoring, debugging, and scaling in past roles.

Showcase collaboration

Be ready to share examples of how you partnered with product owners, researchers, and cross-functional peers to drive projects to completion.

Questions & practice

Merge partitioned event streams into one ordered feed with bounded lateness

Hard · k-way merge · watermarks · out-of-order streams

The read-model service consumes 64 log partitions carrying about 4,000 events per second in total. Each partition is ordered within itself, but partitions drift by up to 30 seconds, and the activity feed must present a tenant's events in occurred_at order. Produce the merge. State its complexity, the buffer it requires in events and in bytes, what happens when one partition is idle, and what you do with an event that arrives after you have already emitted its position. Payloads average 1 KB.

Approach

Merge with a min-heap over the 64 partition heads keyed on (occurred_at, event_id): O(log P) per event and O(n log P) overall. The tie-break on event_id is what makes the output deterministic when two partitions carry the same millisecond, which matters because the feed is paginated and a non-deterministic order reorders pages under the reader.
Emitting the heap head is only correct once every partition has produced everything up to that timestamp, so the emit condition is a watermark: the minimum across partitions of the highest occurred_at seen, less the allowed lateness. Events are held until the watermark passes them, which is what turns individually ordered streams into a jointly ordered one.
Size the buffer from the lateness rather than guessing: 4,000 events per second times 30 seconds is 120,000 buffered events, and at 1 KB each about 120 MB of heap. That number is the real price of the ordering guarantee and belongs in front of whoever asked for it.
Handle the idle partition explicitly, because it fails the feed rather than corrupting it: a partition with no traffic never advances its own maximum, so the watermark freezes and output stops entirely. Either every partition emits a periodic idle marker carrying the broker's current time, or the watermark falls back to wall clock for a partition silent beyond a threshold.

Follow-up

The lateness budget is raised to five minutes. What is the new buffer, and what besides memory changes?
The consumer restarts. Where does it resume from, and what does the feed look like for the first 30 seconds?

Find overlapping job attempts and peak concurrency from lease records

Medium · sweep line · intervals · leases

A day of job_run history yields about 50,000,000 attempt records: (job_run_id, job_type, attempt, started_at, finished_at which is NULL when the worker died, lease_expires_at). Leases expire on a clock, so a job that outran its lease ran twice. Produce (a) every job_run_id whose attempts overlapped in wall-clock time and (b) the peak number of simultaneously running attempts per job_type with the minute it occurred. Target O(n log n). State how you treat a NULL finished_at and what clock skew does to your answer.

Approach

Define the interval before sorting anything: an attempt occupies [started_at, COALESCE(finished_at, lease_expires_at)). finished_at is observed and lease_expires_at is only a promise, so every attempt without a finish contributes an estimate and the whole result is a lower bound on overlap rather than an exact count.
For peak concurrency, sweep: emit 2n endpoints, sort by (timestamp, kind) with ends ordered before starts at equal timestamps, then walk the sequence maintaining a counter per job_type and record each type's maximum with its timestamp. O(n log n) dominated by the sort, O(n) space, or O(1) extra if the sort is external and the walk streams.
For overlap detection, do not compare attempts pairwise. A single global sort by (job_run_id, started_at) gives both the grouping and the order; within a group, keep the maximum end seen so far and report an overlap exactly when the next start is less than that running maximum, which is one linear pass after the sort.
Half-open intervals matter and are easy to get wrong: with closed intervals an attempt ending at the same millisecond another begins reads as concurrency two, and across 50,000,000 records that artefact swamps the real signal.

Follow-up

A handler is not idempotent and you have found 400 overlapping jobs. Which of them actually caused damage, and what would you query to find out?
Peak concurrency for one job_type is 4 against a configured cap of 4. Is the cap working, or is the data hiding attempts that never started?

Collapse a redelivered event batch into per-aggregate high-water marks

Easy · hashing · at-least-once · aggregation

You drain a batch of up to 5,000,000 events, each (aggregate_id BIGINT, aggregate_version INT, event_type, payload). The log guarantees order within one aggregate only; the batch merges 64 partitions, and a relay failover has redelivered a range, so an older version for an aggregate can appear after a newer one. Given a map of last_applied_version per aggregate, produce the events worth applying, at most one per (aggregate_id, version), plus the count discarded. Target O(n) time. State the memory for 2,000,000 distinct aggregates and what you do when it does not fit.

Approach

One pass, one hash map from aggregate_id to the highest version kept, and a discard counter. An event whose version is at or below last_applied_version for its aggregate is dropped without further work, which is the whole reason the event carries its version rather than a delta. O(n) expected time, O(d) space in distinct aggregates.
Keep the maximum, never the last occurrence. The redelivered range means the final appearance of an aggregate in the batch can be an older version than one seen earlier in the same batch, so last-wins applies stale state over newer state and the projection regresses with no error anywhere.
Cost the memory instead of calling it large: an 8-byte key plus a 4-byte version is 12 bytes of payload, and an open-addressed table held at a 0.7 load factor costs roughly 17 bytes per entry before per-slot metadata, so 2,000,000 aggregates is tens of megabytes in a native layout and several times that in a runtime that boxes both key and value.
If the distinct set exceeds memory, partition on hash(aggregate_id) mod P and reduce each partition independently. Every event for one aggregate hashes to the same partition, so the per-partition result is exact and the merge is concatenation rather than a second reduction.

Follow-up

The payload is a patch rather than a snapshot, so applying only the highest version loses the intermediate changes. What changes in your reduction?
How do you detect that version 7 arrived while version 6 was never delivered, and what should the consumer do about the gap?

Explain why the owner filter ignores the listing index

Medium · composite index · query plans · sort elimination

The only index on resource is (tenant_id, status, updated_at DESC, resource_id DESC). A new endpoint returns one user's resources across all statuses, newest created first: WHERE tenant_id = $1 AND owner_user_id = $2 ORDER BY created_at DESC LIMIT 20. On a tenant with 2M rows it takes 900 ms and EXPLAIN shows a sort above a large scan. Explain precisely why the existing index cannot serve it, give the index that can, and state which of these the new index still will not help: owner_user_id alone across tenants; the same query ordered by updated_at. PostgreSQL 16.

Approach

Separate the two jobs an index does. For filtering, a composite btree is seekable only on a left prefix, so with no predicate on status the scan can at best range over tenant_id and test owner_user_id per row; PostgreSQL 16 has no btree skip scan to jump the unconstrained column.
For ordering, the index is sorted by (status, updated_at) within a tenant and not by created_at, so the LIMIT cannot stop early: every matching row is read and then sorted. That is the 'Sort Method: top-N heapsort' line, and it is why the plan reads 2M rows to answer with 20.
Derive the replacement from the access path — equality, equality, then the ordering column: CREATE INDEX CONCURRENTLY ON resource (tenant_id, owner_user_id, created_at DESC). The scan seeks to the (tenant, owner) range and walks 20 entries in order, so the Sort node disappears along with the row-read.
Treat INCLUDE (title, status) as conditional, not free. An index-only scan still visits the heap for any row whose page is not marked all-visible, so on a table taking 1.2k writes/second the win depends on autovacuum keeping the visibility map current, and the wider index costs more on every insert.

Follow-up

90% of rows are status='active'. Would a partial index WHERE status = 'active' change your answer, and for which of the three queries?
A dashboard runs this for 40 owners in one page load. What changes about the design?

Find version gaps and relay lag with window functions

Medium · window functions · event tables · frame semantics

outbox_event holds event_id, aggregate_type, aggregate_id, aggregate_version, event_type, payload, status ('pending','published','dead'), attempts, created_at, published_at. A projection is missing rows and you must decide whether the relay skipped events or the consumer dropped them. Write three queries over the last seven days: one listing every aggregate_id whose published aggregate_version sequence has a hole, one giving per-day counts with a running total, and one returning the newest published event per aggregate. For each, say where the window function is evaluated relative to WHERE and LIMIT. PostgreSQL 16.

Approach

Gaps: compute lead(aggregate_version) OVER (PARTITION BY aggregate_id ORDER BY aggregate_version) in a subquery, then filter next_version <> aggregate_version + 1 in the outer query. Window functions are evaluated after WHERE, GROUP BY and HAVING and before the outer ORDER BY and LIMIT, so the predicate cannot sit in the same WHERE clause and PostgreSQL 16 has no QUALIFY.
Say what the seven-day filter does to the answer: it truncates every partition, so the first row per aggregate has no predecessor inside the window and a hole spanning the boundary is invisible. Widen the window, or join to resource.version as the authority for the true maximum.
Running total: SELECT date_trunc('day', created_at) AS d, count() AS n, sum(count()) OVER (ORDER BY date_trunc('day', created_at) ROWS UNBOUNDED PRECEDING). An aggregate inside a window call is legal because grouping runs before windowing. The grouping key is unique per row here so ROWS and RANGE agree, but write the frame anyway — over ungrouped rows with tied timestamps the default RANGE frame pulls in every peer row and the total jumps.
Newest per aggregate: DISTINCT ON (aggregate_id) ... ORDER BY aggregate_id, aggregate_version DESC is the cheap PostgreSQL-only form when an index matches that order; row_number() OVER (PARTITION BY aggregate_id ORDER BY aggregate_version DESC) = 1 is the portable form and needs a subquery for the same evaluation-order reason as the gap query.

Follow-up

Relay failover redelivers events. Does a duplicate break the gap query, and how would you detect one from this table alone?
Turn the gap check into a continuous monitor rather than a query someone runs after an incident. What does it watch?

Walk me through your most complex backend architecture project and the trade-offs you made.

Medium · Technical Experience and Fundamentals

Approach

Fix the scope first: who calls this, how often, and what they do when it fails.
Name the read and write paths separately; they rarely have the same bottleneck.
Choose a partition key and say what query it makes expensive.
State the consistency you need, and where you are willing to be stale.

Follow-up

What breaks first when traffic grows ten times?
How does this behave when that dependency is down for an hour?

What experience do you have with cloud-based infrastructure and deployment technologies like AWS?

Medium · Technical Experience and Fundamentals

Approach

Clarify what is being asked and what a complete answer contains.
State your assumptions explicitly before working the problem.
Say what you would check first and why it is the highest-information step.
Work from the requirement backwards to the design.

Follow-up

What assumption would you test first?
How would you know your answer was wrong?

How would you design an API that integrates with multiple third-party point-of-sale systems?

Medium · Technical Experience and Fundamentals

Approach

Say who the caller is and what they do when the call fails halfway.
Define the identity of a request so a retry cannot double-apply it.
Separate accepted, pending, failed and confirmed; they are different facts.
Design the error taxonomy before the success shape; callers branch on it.

Follow-up

What happens if the caller retries after a timeout?
How does a client discover it is on an old version of this contract?

What are the primary bottlenecks when scaling backend systems in production environments?

Medium · Technical Experience and Fundamentals

Approach

Clarify what is being asked and what a complete answer contains.
State your assumptions explicitly before working the problem.
Say what you would check first and why it is the highest-information step.
Work from the requirement backwards to the design.

Follow-up

What assumption would you test first?
How would you know your answer was wrong?

How would you design a real-time, menu-aware ordering system that handles ambiguous customer requests?

Medium · System Design and Architecture

Approach

Fix the scope first: who calls this, how often, and what they do when it fails.
Name the read and write paths separately; they rarely have the same bottleneck.
Choose a partition key and say what query it makes expensive.
State the consistency you need, and where you are willing to be stale.

Follow-up

What breaks first when traffic grows ten times?
How does this behave when that dependency is down for an hour?

How do you approach zero-to-one backend system design for a new product initiative?

Medium · System Design and Architecture

Approach

Fix the scope first: who calls this, how often, and what they do when it fails.
Name the read and write paths separately; they rarely have the same bottleneck.
Choose a partition key and say what query it makes expensive.
State the consistency you need, and where you are willing to be stale.

Follow-up

What breaks first when traffic grows ten times?
How does this behave when that dependency is down for an hour?

How would you architect a backend to reliably ingest and process audio streams in noisy environments?

Medium · System Design and Architecture

Approach

Fix the scope first: who calls this, how often, and what they do when it fails.
Name the read and write paths separately; they rarely have the same bottleneck.
Choose a partition key and say what query it makes expensive.
State the consistency you need, and where you are willing to be stale.

Follow-up

What breaks first when traffic grows ten times?
How does this behave when that dependency is down for an hour?

What containerization strategies using Docker and Kubernetes would you employ for a microservices architecture

Medium · System Design and Architecture

What containerization strategies using Docker and Kubernetes would you employ for a microservices architecture?

Approach

Fix the scope first: who calls this, how often, and what they do when it fails.
Name the read and write paths separately; they rarely have the same bottleneck.
Choose a partition key and say what query it makes expensive.
State the consistency you need, and where you are willing to be stale.

Follow-up

What breaks first when traffic grows ten times?
How does this behave when that dependency is down for an hour?

How do you monitor and optimize backend performance metrics in production?

Medium · System Design and Architecture

Approach

Fix the scope first: who calls this, how often, and what they do when it fails.
Name the read and write paths separately; they rarely have the same bottleneck.
Choose a partition key and say what query it makes expensive.
State the consistency you need, and where you are willing to be stale.

Follow-up

What breaks first when traffic grows ten times?
How does this behave when that dependency is down for an hour?

One customer endpoint stalls deliveries to every other destination

Medium · bulkhead · timeouts · saturation

The egress service delivers about 1.5k webhooks/second across 40,000 destinations, with a per-destination concurrency cap of 4 and a 10-second connect-plus-read timeout. Throughput falls to 300/second, queue depth climbs, and p99 delivery latency for unaffected destinations goes from 200 ms to minutes, while the error rate barely moves. One tenant holds 900 destination rows whose URLs share a hostname that now answers in 9.5 seconds. Explain the mechanism with the arithmetic, then give the containment in the order you would apply it.

Approach

Look at saturation before errors. A flat error rate with collapsing throughput says nothing is failing, things are waiting, so the first signal to pull is in-flight request count or pool wait time rather than the error counter. This is the distinction that decides the whole investigation.
Group in-flight work by resolved host, not by destination id. The cap is keyed per destination row, so 900 rows sharing one hostname buy 3,600 concurrent slots against a single host, each held for 9.5 seconds. The bulkhead was never a bulkhead for that host, and grouping by the wrong dimension is why the dashboard looked healthy.
Do the arithmetic in both directions. Required concurrency is arrival rate times latency, so 1.5k/second at 200 ms needs about 300 in flight, which is entirely consumed by 3,600 slow slots; conversely whatever concurrency is left sustains rate equals concurrency divided by 9.5 seconds, which is the 300/second you are seeing. Matching both numbers is what promotes this from a plausible story to the mechanism.
Explain why the circuit breaker never helped. It opens on consecutive failures, and a 9.5-second response inside a 10-second timeout is a success. Slow is not failing, so an error-rate breaker cannot see this; you need a slow-call ratio, a deadline propagated from the caller's remaining budget, or a concurrency limiter.

Follow-up

The host recovers to 80 ms. How long does the queue take to drain, and what does the drain do to the recovered host?
Where should the 10-second timeout number actually come from?

Suggested practice checklist

Built from the rounds and topics Deepgram candidates report.

Day 1 — Map the Deepgram loop

Write out the reported sequence: Recruiter Screen, Technical Screen, Take-Home Assignment, Virtual Onsite.
For each round, write one sentence on what it is judging, from the description above, and mark the one you are least ready for.

Deliverable: A one-page map of the 4 reported rounds, with the weakest marked.

Day 2 — Work Data Structures Fundamentals

Spend the session on Data Structures Fundamentals, which Deepgram candidates report being tested on.
Write one worked example in Data Structures Fundamentals and time yourself on it.

Deliverable: One timed worked example in Data Structures Fundamentals.

Day 3 — Work Technical Screen (experience-based)

Spend the session on Technical Screen (experience-based), which Deepgram candidates report being tested on.
Write one worked example in Technical Screen (experience-based) and time yourself on it.

Deliverable: One timed worked example in Technical Screen (experience-based).

Day 4 — Work Take-Home Assignment Development

Spend the session on Take-Home Assignment Development, which Deepgram candidates report being tested on.
Write one worked example in Take-Home Assignment Development and time yourself on it.

Deliverable: One timed worked example in Take-Home Assignment Development.

Day 5 — Answer out loud: Technical Experience and Fundamentals

Answer aloud, timed: Walk me through your most complex backend architecture project and the trade-offs you made.
Answer aloud, timed: How do you handle data structure fundamentals when optimizing high-throughput data streams?

Deliverable: Spoken answers to 2 reported Technical Experience and Fundamentals question(s), under time.

Day 6 — Answer out loud: System Design and Architecture

Answer aloud, timed: How would you design a real-time, menu-aware ordering system that handles ambiguous customer requests?
Answer aloud, timed: How do you approach zero-to-one backend system design for a new product initiative?

Deliverable: Spoken answers to 2 reported System Design and Architecture question(s), under time.

Day 7 — Answer out loud: Behavioral and Collaboration

Answer aloud, timed: Tell me about a time you had to deal with ambiguous requirements and how you resolved them.
Answer aloud, timed: How do you prioritize tasks when working across multiple projects in a fast-paced startup environment?

Deliverable: Spoken answers to 2 reported Behavioral and Collaboration question(s), under time.

Behavioural preparation

Behavioural rounds judge the decision you made and what it cost.

How do you handle data structure fundamentals when optimizing high-throughput data streams?

Medium · Technical Experience and Fundamentals

Approach

Pick a story where you made the decision, not one where you watched it.
State the situation in two sentences and spend the rest on the reasoning.
Give the blast radius: what could have broken, and what you measured.
Name the disagreement and how you resolved it with evidence.

Follow-up

What would you do differently if you ran that again?
How did you know your change caused the improvement?

Tell me about a time you had to deal with ambiguous requirements and how you resolved them.

Medium · Behavioral and Collaboration

Approach

Pick a story where you made the decision, not one where you watched it.
State the situation in two sentences and spend the rest on the reasoning.
Give the blast radius: what could have broken, and what you measured.
Name the disagreement and how you resolved it with evidence.

Follow-up

What would you do differently if you ran that again?
How did you know your change caused the improvement?

How do you prioritize tasks when working across multiple projects in a fast-paced startup environment?

Medium · Behavioral and Collaboration

Approach

Pick a story where you made the decision, not one where you watched it.
State the situation in two sentences and spend the rest on the reasoning.
Give the blast radius: what could have broken, and what you measured.
Name the disagreement and how you resolved it with evidence.

Follow-up

What would you do differently if you ran that again?
How did you know your change caused the improvement?

Describe a situation where you had to collaborate closely with machine learning engineers or product teams.

Medium · Behavioral and Collaboration

Approach

Pick a story where you made the decision, not one where you watched it.
State the situation in two sentences and spend the rest on the reasoning.
Give the blast radius: what could have broken, and what you measured.
Name the disagreement and how you resolved it with evidence.

Follow-up

What would you do differently if you ran that again?
How did you know your change caused the improvement?

How do you handle disagreements on technical direction with your peers or management?

Medium · Behavioral and Collaboration

Approach

Pick a story where you made the decision, not one where you watched it.
State the situation in two sentences and spend the rest on the reasoning.
Give the blast radius: what could have broken, and what you measured.
Name the disagreement and how you resolved it with evidence.

Follow-up

What would you do differently if you ran that again?
How did you know your change caused the improvement?