How to Apply to Together AI

Key Takeaways

Study Together AI's technical blog posts, open-source projects (like RedPajama), and published research before applying — reference specific work in your application to demonstrate genuine engagement with their mission.
Tailor your resume to the exact role you're targeting, using the same technical terminology found in the job description (e.g., 'LLM inference optimization,' 'GPU cluster operations,' 'model fine-tuning') to align with Greenhouse keyword searches.
Quantify every major achievement on your resume with performance metrics — latency improvements, throughput gains, cluster scale, cost reductions — because Together AI competes on infrastructure performance and your ability to drive measurable impact is the primary hiring signal.
Prepare for technically rigorous interviews by brushing up on distributed systems design, GPU computing fundamentals, and ML inference serving architectures — expect interviewers who are domain experts and will probe your understanding several layers deep.
Highlight any open-source contributions, research publications, or community involvement in AI/ML — Together AI's culture is rooted in open-source values, and demonstrating that you share this ethos differentiates you from candidates with comparable technical skills.
Move quickly and communicate proactively throughout the process — Together AI is scaling rapidly in a competitive talent market, and responsiveness signals the urgency and ownership mentality they value.

About Together AI

Together AI is a leading AI infrastructure company building a cloud platform that enables developers, researchers, and enterprises to train, fine-tune, and run open-source large language models at scale. Founded by a team of prominent AI researchers — including contributors to foundational work like FlashAttention and the RedPajama dataset — the company has rapidly established itself as a go-to alternative to closed-model providers, championing the open-source AI ecosystem. Together AI's platform offers inference, fine-tuning, and custom model services optimized for performance and cost-efficiency, directly competing with hyperscalers and proprietary AI providers. What draws top talent to Together AI is the rare intersection of cutting-edge research and real-world product impact. Engineers and researchers here don't just publish papers — they ship infrastructure that thousands of developers rely on daily. The company's culture reflects its research roots: intellectual rigor, deep technical discourse, and a bias toward first-principles thinking. With roles spanning GPU cluster operations, ML inference optimization, hardware sourcing, and developer experience, Together AI operates more like a vertically integrated AI systems company than a typical SaaS startup. The team is globally distributed with a strong San Francisco presence, and the pace is decidedly startup-fast. If you thrive in environments where you're solving problems at the frontier of systems engineering and machine learning — optimizing inference kernels, managing large-scale GPU fleets, or designing APIs that abstract enormous complexity — Together AI offers an unusually compelling opportunity. The company has attracted significant venture funding and is scaling aggressively, making this a high-growth moment for new hires to shape both the product and the culture.

Application Process

Identify Your Best-Fit Role on Together AI's Careers Page
Visit Together AI's careers page (linked from together.ai/about) to browse their roughly 42 open positions across infrastructure, ML engineering, hardware operations, design, and developer experience. Pay close attention to the specific domain indicated in each title — roles like 'Customer Support Engineer (GPU Cluster)' and 'LLM Inference Frameworks and Optimization Engineer' signal deeply specialized positions, not generalist openings. Identify where your specific technical depth aligns rather than applying broadly.
Research Together AI's Technical Stack and Open-Source Contributions
Before applying, explore Together AI's technical blog, GitHub repositories, and published research papers from team members. Understanding projects like RedPajama, their inference optimization approaches, and their API architecture will help you tailor your application with relevant context. This research is especially critical because Together AI's hiring culture values candidates who demonstrate genuine intellectual engagement with the company's technical mission.
Submit Your Application Through Greenhouse
Together AI uses Greenhouse as their applicant tracking system, so all applications flow through structured job posting forms. Complete every field carefully — Greenhouse allows recruiters to filter and search by specific keywords, so incomplete profiles get deprioritized. Upload your resume as a clean PDF, fill in any supplemental questions thoroughly, and include a brief note in any open-text fields explaining your specific interest in Together AI's mission around open-source AI infrastructure.
Initial Recruiter or Hiring Manager Screen
For a startup of Together AI's profile, initial screens are commonly conducted by a recruiter or directly by the hiring manager, especially for senior technical roles. Expect this call to focus on your background, your understanding of Together AI's platform and market position, and your motivation for joining an AI infrastructure startup. Be prepared to discuss specific technical projects from your past work in concrete detail — vague summaries won't pass muster at a company where deep expertise is the baseline.
Technical Assessment or Take-Home Challenge
Depending on the role, Together AI typically includes a technical evaluation — this could be a live coding session focused on systems-level problems, an ML-specific assessment involving model optimization or inference benchmarking, or a take-home project. For infrastructure roles, expect questions around distributed systems, GPU computing, and low-level performance optimization. For ML roles, anticipate problems involving transformer architectures, quantization, or serving frameworks.
On-Site or Virtual Technical Deep-Dive Interviews
The core interview loop at an AI infrastructure startup like Together AI commonly involves 3-5 sessions with engineers, researchers, and cross-functional leads. Expect a mix of systems design (e.g., designing a distributed inference pipeline), coding (often in Python or C++/CUDA), and domain-specific discussions. For non-engineering roles like the Global Hardware Sourcing Manager or Lead Product Designer, expect portfolio reviews, case studies, or operational scenario discussions relevant to the role's domain.
Final Conversations and Offer Stage
Final-round conversations at Together AI may include a discussion with a co-founder or senior leadership, particularly for senior or strategically important roles. This stage typically assesses culture alignment, long-term thinking, and your vision for the role's impact. Offers from well-funded startups like Together AI commonly include competitive base compensation alongside meaningful equity, reflecting the high-growth trajectory of the company.

Resume Tips for Together AI

Critical Lead With AI Infrastructure and Systems-Level Experience

Together AI's roles center on large-scale AI systems — GPU cluster management, inference optimization, distributed training, and cloud platform engineering. Your resume's top third should immediately surface experience with these domains. Instead of generic bullet points like 'Worked on backend services,' write 'Designed and operated a multi-node GPU inference pipeline serving 10K+ requests/second with P99 latency under 200ms.' Together AI's hiring teams scan for signal that you've operated at the intersection of ML and systems engineering.

Critical Incorporate Together AI's Specific Technical Vocabulary

Greenhouse's search and filter functionality means recruiters can surface candidates by keyword. Use terminology that mirrors Together AI's own language: 'LLM inference,' 'model fine-tuning,' 'open-source models,' 'GPU clusters,' 'CUDA optimization,' 'transformer serving,' 'vLLM,' 'FlashAttention,' 'quantization (GPTQ, AWQ),' and 'API platform.' If you've worked with frameworks or techniques Together AI references in their blog or job descriptions, name them explicitly rather than using generic equivalents.

Critical Quantify Performance and Scale Metrics Relentlessly

AI infrastructure is a domain where performance numbers matter enormously — throughput, latency, cost-per-token, cluster utilization rates, and uptime SLAs. Every bullet point on your resume should aim to include a quantified outcome. Together AI is building products that compete on speed and cost-efficiency, so demonstrating that you've personally driven measurable improvements in these dimensions is the strongest possible signal. A bullet like 'Reduced inference latency by 40% through kernel-level optimization and batching strategies' speaks their language.

Showcase Open-Source Contributions and Research Output

Together AI was built by researchers who value open-source contribution and intellectual output. If you've contributed to open-source ML frameworks (PyTorch, Hugging Face Transformers, vLLM, Triton), published papers, or released datasets, create a dedicated section on your resume for these. Include GitHub links, paper titles with venue names, and download/star counts where impressive. Even a well-regarded blog post on inference optimization or CUDA programming can differentiate you.

Tailor Your Resume to the Specific Role's Domain

Together AI's 42 open roles span remarkably different domains — from hardware supply chain management to documentation engineering to ML research. A resume optimized for the 'LLM Inference Frameworks and Optimization Engineer' role should emphasize CUDA, compiler optimization, and low-level systems work, while an application for the 'Lead DX Engineer - Documentation' role should foreground technical writing, developer empathy, and API documentation experience. Don't submit a one-size-fits-all resume; Greenhouse makes it easy for recruiters to see if your experience aligns with the specific job's requirements.

Format for Clean Greenhouse Parsing

Greenhouse parses resumes to auto-populate candidate profiles, and parsing errors create friction. Use a single-column layout, standard section headers (Experience, Education, Skills, Projects), and avoid tables, multi-column formats, headers/footers with critical information, or heavy graphic elements. Submit as a PDF to preserve formatting. Ensure your name, email, phone, and LinkedIn URL appear at the top of the document in plain text, not embedded in an image or styled text box.

Include Startup and High-Growth Environment Experience

Together AI is a fast-scaling startup, and hiring managers evaluate whether candidates can thrive in that context. If you've worked at startups or high-growth companies, highlight experiences where you wore multiple hats, shipped under tight timelines, or made architectural decisions with incomplete information. Phrases like 'founding engineer,' 'built from zero to production,' or 'scaled system during 10x user growth' signal startup readiness without requiring you to explicitly state it.

Add a Concise Technical Skills Section With Depth Indicators

Rather than a flat list of technologies, organize your skills section to reflect depth: separate 'Expert' from 'Proficient' from 'Familiar.' For Together AI roles, prioritize skills like Python, C++, CUDA, PyTorch, Kubernetes, Terraform, distributed systems, and specific ML serving frameworks. This helps Greenhouse keyword matching while also giving human reviewers a quick depth assessment.

ATS System: Greenhouse

Greenhouse is a structured hiring platform widely adopted by high-growth startups and tech companies. It organizes candidates into scorecarded pipelines, enabling Together AI's hiring team to evaluate applicants consistently across defined criteria. Greenhouse parses uploaded resumes to populate candidate profiles, but also relies heavily on structured application fields, so completing every form element is essential.

Submit your resume as a standard PDF with a single-column layout — Greenhouse's parser handles this format most reliably, avoiding misattributed sections or dropped content.
Mirror exact keywords from the job description in your resume and application responses. Greenhouse allows recruiters to search candidate pools by keyword, and Together AI's roles use specific terminology like 'LLM inference,' 'GPU clusters,' and 'model fine-tuning' that you should echo precisely.
Complete every field in the application form, including optional ones. Greenhouse tracks profile completeness, and incomplete applications may be deprioritized in recruiter searches.
Avoid using tables, text boxes, columns, or infographic-style formatting in your resume. Greenhouse's parser can misread these elements, scrambling your work history or dropping entire sections.
Use standard section headers — 'Work Experience,' 'Education,' 'Skills,' 'Projects' — rather than creative alternatives. Greenhouse maps content to structured fields based on header recognition.
If the application includes open-text fields or 'Additional Information' sections, use them strategically to explain your specific interest in Together AI and how your experience maps to their open-source AI infrastructure mission.
Keep your file name professional and identifiable (e.g., 'FirstName_LastName_Resume.pdf') — recruiters downloading from Greenhouse see the original filename.

Complete Greenhouse Resume Guide

Interview Culture

Together AI's interview process reflects its identity as a research-driven AI infrastructure company staffed by deeply technical builders. Expect a process that values intellectual depth over polish — interviewers are more interested in how you reason through hard problems than in rehearsed answers. The typical interview loop spans 3-5 rounds after the initial screen. For engineering and ML roles, this commonly includes a systems design session where you might architect a distributed inference system, a coding interview emphasizing performance-sensitive programming (Python, C++, or CUDA depending on the role), and a domain deep-dive where you discuss past projects in granular technical detail. Interviewers at Together AI are often themselves researchers or senior engineers who contributed to widely-cited work, so surface-level answers get probed quickly. Be prepared to go several layers deep on any claim you make about your experience. For non-engineering roles — design, developer experience, hardware operations — expect portfolio reviews, case studies, or operational scenario exercises. The 'Global Hardware Sourcing & Supply Manager' role, for instance, would likely involve discussions around GPU supply chain logistics, vendor relationships, and capacity planning under uncertainty. Culture fit at Together AI centers on a few key signals: genuine passion for open-source AI and making powerful models accessible; comfort with ambiguity and the pace of a well-funded but fast-moving startup; and a collaborative, low-ego approach to problem-solving. The company's research roots mean intellectual humility matters — being able to say 'I don't know, but here's how I'd figure it out' is valued over bluffing. Many applicants report that Together AI's process moves relatively quickly compared to larger tech companies, reflecting the startup's urgency to hire exceptional talent in a competitive AI labor market. Responsiveness and enthusiasm throughout the process — following up promptly, asking thoughtful questions, and demonstrating genuine engagement with the company's mission — can meaningfully influence outcomes.

What Together AI Looks For

Deep systems-level engineering expertise — experience building and operating infrastructure at the intersection of distributed computing and GPU-accelerated workloads, not just application-layer development
Genuine engagement with the open-source AI ecosystem — contributions to open-source projects, published research, or demonstrated fluency with open models and the community around them
Performance optimization mindset — a track record of measurably improving latency, throughput, cost-efficiency, or reliability in production systems, ideally in ML-serving or high-performance computing contexts
Startup adaptability and ownership mentality — comfort operating with high autonomy, making consequential technical decisions without extensive process, and iterating rapidly in a scaling environment
First-principles technical reasoning — the ability to break down novel problems (e.g., optimizing inference for a new model architecture) rather than relying solely on established playbooks
Collaborative, low-ego communication style — Together AI's team includes world-class researchers and engineers, and effective collaboration across domains requires intellectual humility and clear communication
Customer and developer empathy — understanding that Together AI's platform serves developers and enterprises, so building reliable, well-documented, and intuitive products is a core value, not an afterthought

Frequently Asked Questions

How long does the Together AI application-to-offer process typically take?

Based on patterns common at well-funded AI startups, Together AI's process likely moves faster than at large tech companies — many applicants at similar companies report 2-4 weeks from application to offer for strong candidates. However, highly specialized roles like the LLM Inference Frameworks Engineer or Global Hardware Sourcing Manager may involve additional evaluation steps or reference checks that extend the timeline. Applying early when a role is freshly posted, responding promptly to scheduling requests, and being available for interviews on short notice can all help accelerate your candidacy.

Should I submit a cover letter when applying to Together AI?

While Greenhouse may not require a cover letter for every role, including one — especially in the 'Additional Information' field — is a smart differentiator for Together AI. Use it to explain specifically why you're drawn to open-source AI infrastructure, reference a Together AI project or paper that resonated with you, and concisely map your experience to the role's key requirements. Keep it to 200-300 words; Together AI's team is engineering-heavy and values conciseness. A cover letter that demonstrates genuine technical engagement with the company's work is far more impactful than a generic one, and at a startup with a large applicant pool, it's one of the few opportunities to convey personality and motivation beyond your resume.

What level of experience does Together AI expect for its engineering roles?

Together AI's job titles range from individual contributor roles (Machine Learning Engineer, Infrastructure Engineer) to senior and leadership positions (Engineering Manager / Tech Lead, Lead Product Designer), suggesting they hire across experience levels. However, given the company's focus on cutting-edge AI infrastructure, even mid-level roles likely expect candidates with meaningful hands-on experience — not just familiarity — with GPU computing, distributed systems, or ML model serving. If you're earlier in your career, emphasize relevant project work, open-source contributions, or research experience that demonstrates depth in Together AI's core technical domains rather than breadth across unrelated areas.

Does Together AI offer remote work opportunities?

Several of Together AI's job postings indicate geographic flexibility — the 'Customer Support Engineer (GPU Cluster), India' role explicitly references an international location, and many AI startups of this profile support distributed teams. That said, some roles may prefer or require proximity to their San Francisco headquarters, especially those involving close cross-functional collaboration or hardware-related work. Check the specific job posting for location requirements, and if a role doesn't specify, ask during the recruiter screen. Demonstrating experience working effectively in distributed or asynchronous teams strengthens your candidacy for remote positions.

How should I prepare for a technical interview at Together AI?

Preparation should be targeted to Together AI's specific domain. For infrastructure roles, practice designing distributed systems with GPU-specific constraints — think about multi-node inference serving, load balancing across heterogeneous hardware, and fault tolerance in GPU clusters. For ML roles, review transformer architecture internals, inference optimization techniques (quantization, batching strategies, KV-cache optimization), and serving frameworks like vLLM or Triton Inference Server. Go beyond LeetCode-style problems; Together AI's interviewers are more likely to present open-ended systems challenges that reflect real problems they face. Reviewing Together AI's blog and any public talks by team members gives you insight into the types of problems they prioritize.

What makes a strong candidate for Together AI compared to other AI companies?

Together AI sits at the infrastructure layer of the AI stack, which means they value systems thinking and performance engineering more heavily than pure ML research experience. A strong Together AI candidate combines ML fluency with deep infrastructure expertise — someone who understands not just how a transformer model works, but how to serve it efficiently at scale across a GPU cluster. Alignment with open-source values is another major differentiator; candidates who have contributed to open-source ML projects, released tools or datasets, or actively participate in the open AI community signal cultural fit in a way that credentials alone cannot. Finally, startup resilience matters — Together AI needs people who can ship quickly, make pragmatic tradeoffs, and operate with high autonomy.

How do I optimize my application for Together AI's Greenhouse ATS?

Greenhouse processes your resume through a parser that maps content to structured candidate profile fields, so formatting matters. Use a clean, single-column PDF with standard section headers and no tables or graphic elements. More importantly, Greenhouse enables recruiters to search and filter candidates by keywords — so your resume must contain the specific technical terms used in the job posting. If the role mentions 'CUDA optimization,' don't just write 'GPU programming.' If it references 'LLM inference,' use that exact phrase. Complete all application form fields, including optional ones, as Greenhouse tracks completeness and recruiters may filter for fully completed profiles.

Is it appropriate to follow up after submitting my application to Together AI?

A thoughtful follow-up can help, but timing and channel matter. Wait at least 7-10 business days after applying before reaching out, as startups often batch-review applications. LinkedIn is typically the most effective channel — identify the recruiter or hiring manager for your role's team and send a brief, specific message referencing the role you applied for and one concrete reason you're excited about Together AI's work. Avoid generic 'just following up' messages. If you have a mutual connection with someone at Together AI, a warm referral is significantly more effective than a cold follow-up; employee referrals are typically flagged and prioritized in Greenhouse.

What programming languages and technical skills are most important for Together AI roles?

The specific requirements vary by role, but patterns across Together AI's postings suggest a core technical stack. Python is foundational across nearly all engineering and ML roles. C++ and CUDA are critical for inference optimization and systems-level roles — the 'LLM Inference Frameworks and Optimization Engineer' role almost certainly requires deep CUDA proficiency. Familiarity with PyTorch is expected for ML positions, and infrastructure roles likely require experience with Kubernetes, cloud platforms (AWS, GCP), and infrastructure-as-code tools like Terraform. Knowledge of ML serving frameworks (vLLM, Triton, TensorRT) and optimization techniques (quantization, operator fusion, memory-efficient attention) is highly valued for Together AI's core technical roles.

Sample Open Positions

Sources

Together AI - About Us & Careers — Together AI
Together AI Blog - Technical Articles and Company Updates — Together AI
Greenhouse Help Center - How Greenhouse Parses Resumes — Greenhouse Software
Together AI Company Reviews and Interview Insights — Glassdoor