Job intelligence
Research Engineer, Machine Learning Systems
Deepgram
Before relying on the listing: Confirm Application closing time with the employer or source. See verification questions
Executive opportunity brief
The decision in one screen
A compact evidence-first read of compensation, flexibility, mobility and what still needs verification. No universal quality score is used.
Fast read
Opportunity DNA
The available details do not point to a single standout advantage yet. The clearest issue to weigh is location rules could change the fit.
Current comparable position is around the 62th percentile.
Geographic eligibility is restricted to a country or region.
The listing does not make the minimum education or experience threshold clear enough yet.
Sponsorship and relocation support are not clear from the current listing.
Part of the application route can be assessed, but the full external process may contain additional steps.
At least one recent enabled source can still renew this vacancy.
This fingerprint summarizes what is clear, constrained or still unanswered in the current listing. It is a reading aid, not an overall job score.
Decision evidence check
4 of 5 core facts stated
This checks whether the listing states the facts that commonly change an employment decision. It measures disclosure completeness, not whether the opportunity is good or bad.
Facts already stated
- Compensation
- Work arrangement
- Location eligibility
- Employment type
What to verify next
- Application closing time: Is there a firm application deadline, a rolling review window or no stated closing date?
Unknown information is never converted into a negative score. The questions above exist to prevent missing evidence from being mistaken for a favorable term.
Decision levers
What could make or break the fit?
Reasons to look closer
Nothing stands out strongly enough yet to call a clear advantage. The role may still be attractive for reasons the listing does not spell out.
Trade-offs to weigh
- Location rules could change the fit: Geographic eligibility is restricted to a country or region.
Questions worth asking before you invest more time
- Confirm whether a degree is actually required, preferred, or replaceable with equivalent experience.
- Confirm the actual minimum experience requirement rather than assuming preferred background is mandatory.
- Ask whether visa sponsorship is available before investing significant time in the application.
These questions come from details the listing leaves open. An unanswered point is not treated as a disadvantage.
Opportunity neighborhood
Keep your options open
These live roles sit close enough to this opportunity to make useful comparison points. Look across several employers before deciding what pay, flexibility or access trade-offs you are willing to accept.
Sr. Data Scientist
hims & hers · USA
Why it is nearby: Same seniority band · Same country · Similar work arrangement
About 18% lower advertised midpoint
Senior Data Analyst (MedTech)
YipitData · USA
Why it is nearby: Same seniority band · Same country · Similar work arrangement
About 6% lower advertised midpoint
Senior Data Scientist, Product
RevenueCat · EMEA, LATAM, Canada, USA
Why it is nearby: Same seniority band · Same country · Similar work arrangement
About 10% higher advertised midpoint
Data Engineer (L5) - Ads
Netflix · Remote — Worldwide
Why it is nearby: Same seniority band · Similar work arrangement · Same employment type
About 148% higher advertised midpoint
This neighborhood is drawn from current inventory in the same broad occupation. Similarity is based on explicit listing facts; missing details do not count against a job.
Role market context
How much current choice is visible around this kind of work?
This compares the vacancy with current High Paid Job inventory in the broad Data & AI roles family. It describes visible choice and disclosure, not future demand or career quality.
200 most recent current roles sampled for the disclosure measures below.
Distinct employers in the bounded current sample, not an estimate of all employers hiring in the market.
Current sampled roles stating at least four of the five Core Decision Facts.
Requires a numeric amount, explicit ISO currency and supported pay period.
Listings with explicit enough remote, hybrid or on-site evidence to classify.
Missing deadline evidence remains unknown and is not interpreted as an open-ended application window.
Currencies with explicit pay evidence in the sample: USD (44) · EUR (15) · GBP (6) · CZK (1). Raw amounts across different currencies are not compared.
This describes current visible High Paid Job inventory in the same broad occupation family. Percentages use a bounded recent sample and are not a forecast, a complete labor-market census or an employer-quality measure.
Constraint opportunity lab
What changes the size of the current opportunity pool?
Each line applies one stricter evidence-based constraint to the same bounded current role-family sample. It helps you see the trade-off in visible choice without deciding which trade-off is right for you.
Counts only affirmative or possible sponsorship evidence, or relocation explicitly offered. Unknown is not treated as no.
A much stricter subset: only listings with explicit worldwide-remote evidence remain.
Shows the subset with enough stated evidence for a more informed first comparison.
Counts roles explicitly classified as worldwide remote, region-restricted remote or remote with restrictions still unclear.
Requires amount, explicit ISO currency and a supported pay period. Missing pay evidence remains unknown.
Only same-currency opportunities with usable annualized pay enter this comparison. Other currencies are not converted or treated as lower.
Uses only roles where a country is known. Roles without country evidence are excluded from this denominator.
Uses only roles with a stated employment type.
These are counterfactual filters over a bounded sample of current HPJ inventory in the same broad occupation family. They show how visible choice changes when one evidence-based constraint is applied; they do not tell you which constraint you should choose or predict hiring outcomes. The calculation uses the 200 most recent current roles from 1197 visible roles in this broad family.
Salary transparency context
Is pay disclosure typical for this kind of role?
This listing states usable structured compensation. About 34% of the bounded current role sample does the same.
Compensation counts only when amount, ISO currency and supported pay period are all known.
Pay disclosure is limited in the current role sample.
This employer’s current pay-disclosure rate is broadly similar to the current same-role sample.
This compares salary-disclosure behavior in current visible listings. It does not compare raw salary amounts across currencies, infer undisclosed pay as zero, or rate employer quality.
Private offer & negotiation lab
Turn an offer into evidence, not guesswork
Enter an annual base offer and, if you want, your own target. HPJ compares only within one explicit currency and never chooses a target for you.
Negotiation evidence available: advertised pay range, current same-currency market context, strong decision evidence.
Enter an annual base offer to compare it with explicit same-currency evidence. HPJ will not suggest a target for you.
Review the full package · 10 terms
Mark what is confirmed, what needs clarification and what you want to prioritize. These choices and notes stay in this browser.
Usable advertised compensation is available.
Variable compensation is mentioned; verify amount, formula, target and eligibility.
Equity or long-term incentive language is mentioned; verify grant type, vesting and valuation basis.
Retirement or pension benefits are mentioned; verify employer contribution and vesting.
Health-benefit language is mentioned; verify coverage, dependants and employee cost.
Paid-leave language is mentioned; verify the exact entitlement and carry-over rules.
A work-arrangement signal is available; verify schedule details before relying on it.
Relocation support not established
Sponsorship not established
A review cycle is mentioned; verify timing and whether compensation can change at that review.
Questions worth taking into the offer conversation
- Is there a bonus, commission or other variable-pay component, and how is it calculated?
- Does the package include equity or another long-term incentive, and what are the vesting terms?
- What retirement or pension contribution does the employer provide?
- How much paid leave is provided and what are the carry-over or blackout rules?
- If relocation is relevant, what expenses, timing and repayment conditions are covered?
- What exact sponsorship or work-authorization conditions apply to this role?
- When is the first performance and compensation review after joining?
- Is there a firm application deadline, a rolling review window or no stated closing date?
Occupation intelligence
What broader career references connect to this role?
HPJ uses broad comparison families for current-market analysis. External taxonomy results are reference candidates for that broad family, not a declaration that this vacancy has been authoritatively classified into a specific occupation.
European occupation references
This service uses the ESCO classification of the European Commission.
U.S. occupation references
- Data Scientists 15-2051.00
- Computer and Information Research Scientists 15-1221.00
- Social Science Research Assistants 19-4061.00
- Atmospheric and Space Scientists 19-2021.00
- Bioinformatics Scientists 19-1029.01
This site incorporates information from O*NET Web Services by the U.S. Department of Labor, Employment and Training Administration (USDOL/ETA). O*NET® is a trademark of USDOL/ETA.
Career reference research
Skills worth investigating
Official occupational context, paired with exact phrases visible in this vacancy. It is not a skills test or an applicant-fit score.
ESCO · European occupation
data scientist
Usually essential in this ESCO occupation
- statistical modeling techniques
- perform project management
- query languages
- online analytical processing
- develop data processing applications
- data engineeringPhrase in listing
- manage data collection systems
Context-dependent ESCO skills
- manage ICT data architecture
- SPARQL
- use spreadsheets software
- computational biology
- state estimation
O*NET · U.S. occupation
Data Scientists
O*NET occupation skills
- Mathematics
- Critical Thinking
- Reading Comprehension
- Complex Problem Solving
- Programming
- Active Learning
- SpeakingPhrase in listing
- Active Listening
Build my private skill preparation brief Your evidence · Practice · Interview questions
Plan how you might demonstrate or investigate occupational skills. These are broad-family reference examples, not confirmed vacancy requirements. Nothing you enter is sent to High Paid Job.
Private to this browser. Up to 20 reference skills; not a qualification assessment or hiring prediction.
Useful next question: Which of these occupational reference skills does the hiring manager actually expect for this specific role, and what evidence would demonstrate them?
These skills belong to one reference occupation selected for HPJ’s broad career family, not a verified classification of this vacancy. Text overlap only means a phrase appears in the available listing text; it does not establish an employer requirement, your proficiency, or a qualification gap. A missing mention is unknown, not a deficit.
Adjacent current roles to investigate
Senior Android Engineer — Kotlin/Jetpack Compose | Social, Messaging & Video
Travoom · Remote — Worldwide
same seniority band · same employment type · similar work-arrangement evidenceSenior AI Developer - Product Operations
Workleap · Remote — Worldwide
same seniority band · same employment type · similar work-arrangement evidenceSenior .NET Developer
Imaginary Cloud · Remote — Worldwide
same seniority band · same employment type · similar work-arrangement evidenceAdjacent-family roles are exploration prompts from current HPJ inventory. They are not inferred qualification matches and they do not imply a promotion path.
Career move map
What sits one step away?
This listing is currently read as Senior / lead. Nearby roles one visible seniority step away can help you judge whether to widen the search, stretch upward or compare the responsibilities more carefully.
Broader-entry direction
Data Scientist: Product
Thatch · Remote — Worldwide
USD 160,000–230,000 / year Similar work arrangement · Same employment type · Recently posted Advertised midpoint is about 3% lowerData Analyst
Sentra · Remote — Worldwide
Similar work arrangement · Same employment type · Recently postedStretch direction
Sr. Director of Machine Learning
hims & hers · USA
USD 330,000–360,000 / year Same country · Similar work arrangement · Same employment type Advertised midpoint is about 73% higherGTM AI Engineer
Customer.io · LATAM, Canada, USA
USD 140,000–170,000 / year Same country · Similar work arrangement · Same employment type Advertised midpoint is about 23% lowerThese are live roles one visible seniority step away in the same broad occupation family. They can reveal nearby options, but they are not a guaranteed promotion path and do not account for your personal experience.
Skill bridge
What does this role connect to?
Kubernetes is one of the signals this listing shares with skills appearing often across current Data & AI roles opportunities.
Skills named in this listing
Shared with the current career sample
- Kubernetes Seen in 11 sampled listings
- Machine Learning Seen in 45 sampled listings
Additional signals in this listing: Docker
Application edge
- If these skills match your background, prepare one concrete example that shows how you used Docker, Kubernetes, and Machine Learning rather than relying on a keyword-only claim.
If the named skills are genuinely transferable in your own experience, the occupation map also points to adjacent current families worth researching: Software engineering roles and Product roles. This is exploration context, not a qualification match.
Skill signals come from explicit text mentions. A mention is not automatically a requirement, and absence from the current top-skill list does not make a skill rare or unimportant. The current career comparison draws on 120 sampled listings; 115 contain a recognized skill signal.
Application Decision Lab
Turn the listing into a preparation plan
Separate what the listing actually states from what you still need to prepare or verify. Your selections and short notes stay in this browser and are never submitted to High Paid Job.
Docker
This skill or technology is explicitly mentioned in the available listing text. A mention is not automatically a mandatory requirement; prepare a concrete example only if it is relevant to your experience.
Kubernetes
This skill or technology is explicitly mentioned in the available listing text. A mention is not automatically a mandatory requirement; prepare a concrete example only if it is relevant to your experience.
Machine Learning
This skill or technology is explicitly mentioned in the available listing text. A mention is not automatically a mandatory requirement; prepare a concrete example only if it is relevant to your experience.
Questions worth resolving before you rely on the opportunity
- Application closing time: Is there a firm application deadline, a rolling review window or no stated closing date?
- Work authorization / sponsorship: If sponsorship or work authorization matters to you, what exact immigration or employer-sponsorship conditions apply to this role?
- Application process: What interview, assessment, portfolio, reference or screening stages are part of the hiring process?
Decision analysis
What the listing really tells you
Research Engineer, Machine Learning Systems at Deepgram becomes easier to judge once pay has a useful comparison point, the location rules are clear enough to judge, and the visible application effort can be estimated are viewed together. Here is the current read from the information available.
Pay has a useful comparison point
Among 53 current comparable salaries in USD, the normalized annual midpoint is around the 62th percentile. This compares advertised pay in the same currency and does not include every element of total compensation.
The location rules are clear enough to judge
Location flexibility can be assessed here: Remote with some uncertainty or restrictions, with a Remote Freedom reading of 62/100. The most important location constraint is geographic eligibility is restricted to a country or region..
The visible application effort can be estimated
Provider-mediated application path. The external application may contain additional steps that are not visible from the listing itself.
Source continuity adds useful context
The opportunity currently has 1 renewable source path across 1 source classes. Multiple independent paths can strengthen continuity, but they are not a substitute for checking the application destination.
How this analysis is assembled
This read combines the structured details currently available for pay, flexibility, qualifications, mobility, application effort and source continuity. Missing facts are left open for you to verify.
Analysis evidence coverage
How much analysis can the current evidence support?
Uses the explainable High Pay Score within the job’s current occupation/seniority/currency comparison sample.
Measures observable location flexibility and explicit restrictions; it does not determine whether remote work is personally desirable.
The salary is source-provided and normalizable, but the advertised range is broad enough to reduce comparison precision.
The source posting date is today.
At least one enabled source recently observed the vacancy. The source chain is traceable, but this is not independent employer verification.
There is no universal opportunity score because pay, flexibility and access matter differently to different people. Compare Jobs lets you weight the evidence around your own priorities.
Decision checklist
What is supported, uncertain or worth checking?
This checklist separates supported strengths, visible cautions and missing evidence. It is not an overall recommendation or guarantee that the job is suitable for you.
Supported strengths
No standout strength is supported strongly enough to list here. That is not a negative rating.
Visible cautions
- Geographic eligibility is restricted to a country or region.
- This is remote work with a geographic restriction, not a worldwide-remote vacancy.
Still unknown
- Visa sponsorship is not established by the available listing data.
Pay vs. barriers
Pay and barrier context available
The current pay position can be viewed alongside the stated barriers, but the available detail does not support a stronger conclusion.
Comparable disclosed-pay evidence places this job at about 62 / 100 within its current occupation/seniority sample.
1 explicit Remote Freedom restriction signal is present and is shown separately from qualification requirements.
- Education requirement is not established.
- Required prior-experience level is not established.
This section weighs pay context against clearly stated barriers without collapsing unlike requirements into one universal score. Degree, experience, licensing, clearance and location limits remain separate, and missing qualification details stay unknown.
Beyond the headline
Not enough evidence for a hidden-value finding
Potential hidden value appears only when several distinct non-pay advantages are actually supported. Missing requirements or benefits are never counted as positives.
The annualized salary midpoint is at or above about 62% of the current comparable disclosed-salary sample.
Still check: The current application path is provider-mediated rather than directly classified as an employer-authorized or official route.
Hidden Value is an experimental evidence summary, not an investment-style score or guarantee of employability. It only uses explicit current listing evidence and does not infer that an unmentioned qualification is unnecessary.
Compensation intelligence
How competitive is this pay?
Within the current market range
The annualized midpoint is above the current comparable median but remains within the sample’s middle-to-upper range.
The highlighted band is the middle 50% of the current comparable sample after the outlier guard. The dark marker is this job’s annualized salary midpoint.
Why this analysis?
- This score reflects the current advertised-pay distribution for comparable active jobs in the same currency. It measures pay position, not overall job quality.
- Salary confidence: Medium. The salary is source-provided and normalizable, but the advertised range is broad enough to reduce comparison precision.
- Range quality: Broad but usable range.
Employer compensation communication
Pay transparency at Deepgram
Mixed current pay transparency
This score uses the employer’s current visible listings and combines how often pay is disclosed with how useful those ranges are. Bonus, equity and benefit mentions are shown separately and do not raise the score.
This pattern comes from the employer’s current visible vacancies. It describes hiring and pay transparency in those listings, not employee satisfaction or a permanent judgement of the employer.
Current employer intelligence
Deepgram — current opportunity snapshot
Employer Evidence Profile
How often this employer’s current visible sampled vacancies state decision-critical information. These percentages describe evidence coverage, not employer quality.
Current sampled listings with advertised compensation evidence.
Current sampled listings whose work arrangement is clearer than unknown.
Current sampled listings with a supported country signal.
Current sampled listings with an explicit closing time.
Current sampled listings with a usable posting date.
Current sampled listings that explicitly indicate confirmed, possible or unavailable sponsorship evidence.
Current sampled listings with an explicit relocation signal.
Low coverage means the current visible listings state the information less often. It does not prove that the employer lacks the policy, benefit or support.
Common current occupation families: Sales (16) · Software engineering (4) · Legal (3)
Visible seniority mix: Senior / lead (12) · Leadership (18)
Current employment-type mix: Full time (40)
Most visible hiring countries in this sample: United States (17) · United Kingdom (13) · Singapore (6) · Australia (1)
Other current roles from this employer
Account Executive (Melbourne)
Australia
Leadership · Region-restricted remoteSenior Pre-Sales Solutions Engineer (Singapore)
Singapore
Leadership · Remote — restrictions unclearBilling & Analytics Software Engineer
USA
Seniority not clearly specified · Region-restricted remoteUSD 180,000–245,000 / yearML Ops Infrastructure Engineer
USA
Seniority not clearly specified · Region-restricted remoteUSD 160,000–220,000 / yearThese are other current vacancies visible from the same employer, not endorsements or evidence of future hiring plans.
Current disclosed salary medians by currency
- USD 197,500 median · 15 disclosed salaries
Questions worth asking this employer
- What compensation range applies to this role, and which parts of total compensation are fixed versus variable?
- If work authorization matters to you, what sponsorship policy applies to this specific role and location?
- If a move is required, is relocation assistance available for this position?
- What is the expected application and interview timeline if the listing does not state a closing date?
These questions are prompted by missing or thin evidence in the employer’s current visible vacancies. Missing evidence is not treated as a negative judgement.
This employer snapshot uses the employer’s current visible listings. Remote and sponsorship rates use only vacancies with explicit evidence, while missing information stays unknown. Salary medians are shown separately by currency and need at least five disclosed salaries. Zero confirmed sponsorship means none was explicitly found in this sample, not that sponsorship is unavailable.
This snapshot describes the employer’s current visible vacancies—things such as pay disclosure and work arrangement. It is not an employee-satisfaction or hiring-quality rating, and different currencies stay separate.
Work-location intelligence
Remote Freedom
Remote with some uncertainty or restrictions
Why this score?
- The role is remote but the source identifies a country or regional restriction.
Location restrictions detected
- Geographic eligibility is restricted to a country or region.
Remote Freedom measures observable location flexibility only. Unknown work-location information remains unscored. Timezone wording is penalized only when it appears restrictive; general collaboration across time zones is not treated as a restriction. A score does not guarantee legal work eligibility, payroll availability or visa sponsorship.
International applicant intelligence
Visa & relocation signals
Sponsorship status is shown only when the listing provides clear enough wording. Employer history or headquarters location is not used to guess an answer.
The available listing details do not contain a clear enough sponsorship statement.
The available listing details do not contain a clear enough relocation statement.
Tradeoff check
What’s the catch?
These are visible conditions worth checking before treating the headline compensation as the whole story:
- This is remote work with a geographic restriction, not a worldwide-remote vacancy.
Current alternatives
Better-paying alternatives
These current roles show a materially higher annualized salary midpoint in the same broad occupation and currency. Similar seniority, country, work arrangement, employment type and recency are preferred where possible; higher pay alone does not make an alternative a better fit.
Staff Machine Learning Engineer, Recommendation Systems - Quora
Quora · Canada, Ireland, USA
USD 219,989–320,613 / year Region-restricted remoteWhy comparable: Same seniority band · Same country · Similar work-location arrangement
Lead Machine Learning Engineer
Serve Robotics · Canada, USA
USD 225,000–260,000 / year Region-restricted remoteWhy comparable: Same seniority band · Same country · Similar work-location arrangement
Engineering Manager, Machine Learning (Safety)
Discord · USA
USD 272,000–340,000 / year Region-restricted remoteWhy comparable: Same seniority band · Same country · Similar work-location arrangement
Alternatives are drawn from the current inventory rather than presented as application recommendations. A pay difference needs to be meaningful—about 3% or more at the midpoint—and different currencies are not compared.
Application-path intelligence
What can we see about the application process?
Provider-mediated application path
The current application path comes through a publisher or aggregator. The destination may still lead to an employer-authorized hiring system, but that is not clear enough to assume.
Current destination: jobicy.com · HTTPS
No additional required application step is clear in the available listing text. This does not mean the external application is one-click or friction-free.
Application effort is estimated from visible listing and source details. Optional steps are kept separate from mandatory ones, and the external form may contain additional requirements that are not visible here.
Application safety context
Application safety check
Basic application-safety context available
The current application destination has a basic transport-safety signal such as HTTPS. This does not verify the employer, vacancy, recruiter or application process.
Evidence checked: Source provenance · Application destination · Listing wording
Observable transport-safety signals
- The available application destination uses HTTPS.
Before you share money or sensitive information
- Confirm the vacancy on the employer’s official careers site when practical, especially if a recruiter contacts you outside the listed application path.
- Check that recruiter email domains and application destinations are consistent with the employer or an identifiable recruiting provider.
- Do not pay application, processing, training or equipment fees merely to be considered for a job.
- Do not send cryptocurrency, gift cards, online-banking credentials or PINs as part of an application.
- Treat unusually fast offers or interviews conducted only through informal messaging channels as reasons to verify the employer independently.
This is a limited safety screen, not employer verification and not a fraud determination. Never send money, cryptocurrency, passwords or sensitive banking credentials merely because a listing appears on a job platform.
Original listing details
The section below comes from the listing source. The decision sections above use the job’s structured pay, location, qualification, mobility and application details to give you additional context.
Company Overview
Deepgram is the leading platform underpinning the emerging trillion-dollar Voice AI economy, providing real-time APIs for speech-to-text (STT), text-to-speech (TTS), and building production-grade voice agents at scale. More than 200,000 developers and 1,300+ organizations build voice offerings that are ‘Powered by Deepgram’, including Twilio, Cloudflare, Sierra, Decagon, Vapi, Daily, Cresta, Granola, and Jack in the Box. Deepgram’s voice-native foundation models are accessed through cloud APIs or as self-hosted and on-premises software, with unmatched accuracy, low latency, and cost efficiency. Backed by a recent Series C led by leading global investors and strategic partners, Deepgram has processed over 50,000 years of audio and transcribed more than 1 trillion words. There is no organization in the world that understands voice better than Deepgram.
Company Operating Rhythm
At Deepgram, we expect an AI-first mindset—AI use and comfort aren’t optional, they’re core to how we operate, innovate, and measure performance.
Every team member who works at Deepgram is expected to actively use and experiment with advanced AI tools, and even build your own into your everyday work. We measure how effectively AI is applied to deliver results, and consistent, creative use of the latest AI capabilities is key to success here. Candidates should be comfortable adopting new models and modes quickly, integrating AI into their workflows, and continuously pushing the boundaries of what these technologies can do.
Additionally, we move at the pace of AI. Change is rapid, and you can expect your day-to-day work to evolve just as quickly. This may not be the right role if you’re not excited to experiment, adapt, think on your feet, and learn constantly, or if you’re seeking something highly prescriptive with a traditional 9-to-5.
The Opportunity
Voice is the most natural modality for human interaction with machines. However, current sequence modeling paradigms based on jointly scaling model and data cannot deliver voice AI capable of universal human interaction. The challenges are rooted in fundamental data problems posed by audio: real-world audio data is scarce and enormously diverse, spanning a vast space of voices, speaking styles, and acoustic conditions. Even if billions of hours of audio were accessible, its inherent high dimensionality creates computational and storage costs that make training and deployment prohibitively expensive at world scale. We believe that entirely new paradigms for audio AI are needed to overcome these challenges and make voice interaction accessible to everyone.
The Role
Deepgram is seeking a highly skilled and versatile Machine Learning Engineer to join our Research team. As a Member of the Research Staff, you will partner with research scientists to prototype and validate novel modeling ideas, then scale them through robust training systems for speech technologies, internal tooling, and innovative data strategies. You'll work at the intersection of machine learning, data infrastructure, and internal tooling to support our mission of building world-class speech recognition and synthesis systems. On the Research team, you will experiment with new technologies and techniques, while also working on product-focused deliverables, learning from colleagues with a wide range of expertise in AI and machine learning as you go.
Key Responsibilities
-
Scalable Model Training: Architect and manage horizontally scalable systems that dramatically accelerate the end-to-end training lifecycle for Speech-to-Text (STT) and Text-to-Speech (TTS) models. This includes far more than automated training: the role focuses on making model development significantly faster and more efficient through optimized data preparation and management, high-throughput training pipelines, distributed infrastructure, and automated evaluation tooling.
-
Tooling & Accessibility: Design and implement internal UIs and tools that make ML systems and workflows accessible to non-technical stakeholders across the company. These UIs should be designed to provide transparency and flexibility to internally built tooling.
-
Infrastructure & Tools: Oversee and manage training tooling, job orchestration, experiment tracking, and data storage.
The Challenge
We are seeking Members of the Research Staff who:
-
See "unsolved" problems as opportunities to pioneer entirely new approaches
-
Can identify the one critical experiment that will validate or kill an idea in days, not months
-
Have the vision to scale successful proofs-of-concept 100x
-
Are obsessed with using AI to automate and amplify your own impact
If you find yourself energized rather than daunted by these expectations—if you're already thinking about five ideas to try while reading this—you might be the researcher we need. This role demands obsession with the problems, creativity in approach, and relentless drive toward elegant, scalable solutions. The technical challenges are immense, but the potential impact is transformative.
It's Important to Us That You Have
-
Strong experience with the machine learning research pipeline, particularly in STT or related speech domains. This includes experimenting with and evaluating new architectures and modeling approaches, and implementing large-scale training systems.
-
Proficiency with orchestration and infrastructure tools like Kubernetes, Docker, and Prefect.
-
Familiarity with ML lifecycle tools such as MLflow.
-
Experience building internal tools or dashboards for non-technical users.
-
Hands-on experience with data engineering practices for unstructured audio and text data.
-
Comfortable working in cross-functional teams that include researchers, engineers, and product stakeholders.
Why Join Deepgram?
At Deepgram, you’ll help shape the future of human–machine communication. Our research culture prioritizes ownership, experimentation, and real-world impact. As a Member of the Research Staff, you'll be empowered to build tools and systems that accelerate ML research and product deployment at scale.
Notice: We're aware of individuals impersonating Deepgram recruiters. All legitimate Deepgram recruiting communication comes from an @deepgram.com email address. If you've received a message claiming to be Deepgram, please forward it to careers@deepgram.com.
Listing provenance
Where did this job come from?
See where this vacancy came from and whether more than one current source can still support it. Source names help with traceability and attribution; they are not endorsements or guarantees.
At least one enabled source recently observed the vacancy. The source chain is traceable, but this is not independent employer verification.
- Continuity
- One recent renewable source
- Source-class independence
- One renewable source class
- Last source observation
- Observed very recently
This records when High Paid Job last observed the vacancy in source evidence. It confirms source presence, not that the employer edited the vacancy at that time.
- 1 enabled source recently observed this vacancy.
Source confidence is about traceability and continuity. It is separate from application safety and should not be read as an employer reputation rating.
The vacancy is supplied by an external job publisher. Source information helps show where the listing came from; it is not a guarantee about the employer or vacancy.
Listing source: Jobicy. Destination: jobicy.com. The application opens in a new tab so you can keep this report open.