How to Ask for a Reference by Email (Templates and Examples)
By
Liz Fujiwara
•
Mar 5, 2026
Reference checks remain an important step in hiring for AI and ML roles, with most employers requesting 2–3 professional references after onsite interviews or before extending an offer.
Without prepared contacts, candidates can scramble to respond when references are requested on short notice, causing stress and delays.
Email is the preferred way to request references because it is asynchronous, creates a written record, and allows sharing the job description, resume, and portfolio links across time zones.
Companies may use AI to triage applications, but human references are still relied on to validate problem-solving, communication, and collaboration, and Fonzi helps candidates prepare for these requests in advance.
Key Takeaways
Email is now the standard way to request references, especially in fast-moving AI roles with 2–3 week hiring cycles.
Effective emails clearly explain why you are asking, what role you are targeting, and the expected response deadline.
Fonzi is a curated marketplace for AI talent where prepared references and validated experience help candidates stand out.
Choosing the Right Person to Be Your Reference
Selecting the right potential reference is the first step toward securing a strong recommendation. The goal is to find people who can speak with authority about your skills, work ethic, and contributions.
Who makes an ideal reference for AI/ML roles:
Direct supervisors or collaborators who worked closely with you on meaningful projects, such as tech leads, engineering managers, research managers, senior collaborators, or PIs. Examples include your manager from your 2022–2024 ML infra team at a FAANG company, or your PhD advisor who supervised your 2020–2023 LLM alignment research.
Prioritize depth over brand. Avoid choosing big names who barely know your work. A Staff Engineer who shipped three features with you provides more value than a VP who met you twice.
Aim for a mix: one former manager who can speak to performance, one peer collaborator, such as a Staff Engineer on a shared project, who can discuss technical collaboration, and possibly one cross-functional partner, such as a PM or data scientist, who can speak to communication and impact.
For early-career candidates or career switchers, reasonable alternatives include lab supervisors, bootcamp instructors, open-source maintainers, or freelance clients who have reviewed your code and can vouch for your qualifications.
When to Ask for a Reference by Email
Timing your reference request appropriately shows respect for your contacts’ schedules and demonstrates professionalism to potential employers.
Timing guidelines for different scenarios:
Most tech companies request references after onsite interviews or once a candidate is the preferred hire, usually within 24–72 hours of that decision. Be ready.
At the start of an active job search, for example January 2026 if you aim to switch roles by Q2, line up 2–4 potential references so you are not scrambling when requests come.
For academic or research roles, such as Research Scientist at a lab, formal reference letters may be due by a fixed deadline, for example 15 April 2026. Email references 2–3 weeks before to give them adequate time.
Fast-moving startups and AI-first companies might request references earlier, sometimes after a single panel interview. Warn your references that calls may come quickly.
On Fonzi, companies often review candidate profiles and references just before Match Day, so we encourage candidates to confirm references ahead of that event to ensure a smooth process.
How to Structure an Email Asking for a Reference
A well-structured reference request email makes it easy for busy professionals to say yes and prepare effectively.
Key structural elements:
Use a clear, searchable subject line such as “Reference Request – [Your Name] – [Role Title, Company] – Response by [Date].” This helps your contact find the email later.
Follow a logical flow: Warm greeting → brief context and professional relationship reminder → clear ask for a strong reference → role summary and links → timeline and logistics → appreciation and exit.
Keep the email to 200–350 words with short paragraphs and scannable formatting. Busy tech leads and former managers appreciate brevity.
Add 3–5 bullet points summarizing key projects you worked on together, with concrete metrics like latency improvements, cost reductions, or model accuracy gains.
Include attachments or links: 1-page current resume (PDF), job description URL, and optionally your portfolio, GitHub, Google Scholar, or a short project doc on relevant work (e.g., your LLM evaluation framework).
Email Templates: How to Ask for a Reference in Different Situations
The following templates are tailored for AI/ML candidates requesting references from different types of contacts. Each template contains a clear ask, specific timeline, and a short reminder of shared work.
Every email template uses realistic dates, for example interviewing in March–April 2026, and roles such as Senior LLM Engineer, ML Platform Engineer, Research Scientist in NLP. Keep your tone polite and confident, and avoid apologetic or vague language even if the professional relationship is friendly.
Template 1: Asking Your Current or Recent Manager for a Reference
This template works well when moving within the same industry or transitioning from big tech to a startup. It’s common for AI engineers to use a recent manager who has firsthand knowledge of their work.
What to include:
Context like “We worked together on the recommendation ranking team from June 2022 to January 2026” with mention of key outcomes.
Clarification if the job search is confidential (not yet disclosed to your current company). If so, focus on former supervisors instead.
The company name (e.g., “AtlasAI”), job title, and 2–3 bullet points on what the new position emphasizes: production ML, LLM infra, or research.
A concrete deadline: “The recruiter plans to speak with references between March 12–15, 2026; would you be available one of those days?”
Example email:
Subject: Reference Request – Alex Chen – Senior ML Engineer, AtlasAI – Response by March 10
Hi Jordan,
I hope you are doing well. I wanted to reach out because I am in the final stages of interviewing for a Senior ML Engineer role at AtlasAI, and they have asked for references.
We worked closely on the recommendation ranking team from June 2022 to January 2026, and I thought you would be able to speak to my contributions, particularly our work reducing inference latency by 30 percent and shipping the new embedding pipeline.
Would you feel comfortable serving as a strong, positive reference for me?
The hiring team plans to reach out between March 12 and 15, 2026. I have attached my updated resume and the job description link below.
Role focus: Production ML systems, real-time inference, cross-functional collaboration
Job description: [link]
Please let me know if you’d like any additional details or a quick call to discuss. I completely understand if the timing doesn’t work out.
Best regards,
Alex
Template 2: Asking a Former Manager You Haven’t Spoken With in a While
This template is for former managers or tech leads from earlier roles who can still speak well to your growth and skills. Reconnecting requires a personal touch.
What to include:
A brief, personal catch-up line referencing a team milestone or shared project from that time.
A 2–3 sentence summary of what you’ve done since leaving that role, focusing on relevant AI/ML experience to update their mental model of your career.
A clear statement of why you value their perspective (“you saw my transition from IC to leading our 2020 model re-architecture project…”).
Extra clarity about logistics: whether the company will call them, email them, or request a written letter, and the date range you’re targeting.
Example email:
Subject: Reference Request – Maya Patel – Staff LLM Engineer, HorizonAI – Response by April 5
Hi Dr. Thompson,
I hope this message finds you well. I often think back to our work on the 2020 model re-architecture project; it was a formative experience that shaped how I approach ML systems today.
Since leaving ABC Company in 2021, I have been leading LLM infra work at my current role, including deploying production systems for 10B+ parameter models and building evaluation frameworks for multimodal outputs.
I am now interviewing for a Staff LLM Engineer position at HorizonAI, and they have asked for references. Given that you saw my growth from IC to project lead, I thought your perspective would be especially valuable.
Would you be comfortable serving as a strong reference for me? The hiring team will likely reach out via phone during the week of April 7–11, 2026.
I have attached my current resume and the job description. Please let me know if you would prefer a quick call to catch up first; I would love to hear how things are going with you.
Look forward to hearing from you.
Best regards,
Maya
Template 3: Asking a Professor or Research Supervisor for a Reference
This template serves candidates applying to research roles (Research Scientist, Applied Scientist, or PhD/postdoc positions) where formal letters may be needed.
What to include:
Specific collaborations: thesis titles, papers (with publication years, e.g., NeurIPS 2023), or shared projects like benchmark datasets or RL environments.
Whether it’s a confidential reference letter uploaded via an application portal with a deadline, or a casual phone conversation with an industry hiring manager.
A short bullet list of what you hope the letter will emphasize: independent research ability, LLM safety work, or experience with large-scale training.
All relevant links (Google Scholar, arXiv, project pages) and a clear due date, especially if multiple letters are needed.
Example email:
Subject: Reference Letter Request – James Liu – Research Scientist, DeepMind – Due April 15
Dear Professor Martinez,
I hope you are doing well. I am reaching out because I am applying for a Research Scientist position at DeepMind, and the application requires two formal reference letters.
During my PhD (2019–2024), we collaborated closely on LLM alignment research, culminating in our NeurIPS 2023 paper on reward modeling and the RLHF benchmark we released.
Would you be willing to write a strong letter on my behalf? The portal deadline is April 15, 2026, and letters should be uploaded via the link I will forward separately.
If helpful, I would be grateful if the letter could speak to:
My independent research contributions and initiative
Technical depth in LLM safety and evaluation
Experience coordinating large-scale training experiments
I have attached my current resume and Google Scholar profile for your reference. Please let me know if you would like to discuss anything before writing.
Thank you so much for your time and mentorship over the years!
Best regards,
James
Template 4: Asking a Peer, Tech Lead, or Cross-Functional Partner
This scenario covers Staff Engineers, Staff Data Scientists, PMs, or Tech Leads who closely partnered with you but weren’t your direct manager.
What to include:
How you collaborated (e.g., “we co-led the 2023 migration of our feature store to a new infra stack” or “we shipped our LLM-based support assistant together”).
Framing the request as a complement to manager references, focusing on collaboration, communication, and technical leadership rather than formal performance review.
A suggested focus area (e.g., “If you’re comfortable, it’d be helpful if you spoke to how I handled cross-team coordination and incident response.”).
Context that peer references are especially valuable in high-collaboration environments and sometimes requested explicitly by startups and research labs.
Example email:
Subject: Reference Request – Sam Rivera – ML Platform Engineer, ScaleAI – Response by March 20
Hi Taylor,
I hope things are going well with the platform team. I am reaching out because I am interviewing for an ML Platform Engineer role at ScaleAI, and they have specifically asked for a peer or collaborator reference.
We worked closely on the 2023 feature store migration and shipped the LLM-based support assistant together. I thought you would be able to speak to how we collaborated across teams.
Would you feel comfortable serving as a reference? If so, it would be helpful if you could speak to our cross-team coordination and how I handled incident response during the migration rollout.
The recruiter mentioned they would reach out during the week of March 24 to 28. I have attached my resume and the job description for context.
I completely understand if the timing does not work. Please let me know either way.
Best regards,
Sam
What to Include in Your Reference Request Email
Think of your reference request as a “reference packet,” a concise collection of everything a busy contact needs to advocate effectively for you.
Must-have elements:
Company and role details: Who you’re interviewing with (company name), role title, team focus (e.g., LLM platform team, RLHF team), and approximate start date.
Role summary: A 2–3 sentence description of the role’s focus, like building evaluation pipelines for multimodal models or scaling training infra for 10B+ parameter models.
Shared project highlights: 3–4 specific projects the reference saw, including measurable impact (e.g., improved offline AUC by 4%, reduced inference latency by 30%, shipped production LLM serving system).
Contact details and timeline: How and when they’ll be contacted and specific date ranges.
Attachments: Your current resume with dates through 2026, portfolio or GitHub link, and the official job description so the reference can align their comments.
Offer for more context: Close with an offer to provide additional details or a short call if they prefer to talk things through before speaking with the hiring team.
Examples of Strong vs. Weak Reference Request Emails
Side-by-side comparison helps you spot common mistakes and understand what makes a reference request effective. Most weak emails share similar problems: vague asks, missing context, and unclear timelines.
Weak emails use phrases like “Can you be my reference?” without specifying the role or company. They omit deadlines, forcing the reference to guess the urgency. Stronger emails name the position, company, and response window explicitly.
Examples should feel realistic for AI roles, using specific titles such as “Senior LLM Engineer at HorizonAI” and concrete reference windows, for example “by April 5, 2026.” A strong email shows respect for the reference’s time while confidently signaling that you expect to be a serious candidate.
Comparison Table: Weak vs. Strong Reference Request Emails
The following table highlights key differences between ineffective and effective reference request emails:
Element | Weak Example | Stronger Example |
Subject Line | “Quick question” | “Reference Request – Alex Chen – Senior ML Engineer, AtlasAI – Response by March 10” |
Opening | “Hey, hope you’re doing okay.” | “Hi Jordan, I hope you’re doing well! I wanted to reach out because I’m in final interviews for a role I’m excited about.” |
Clarity of Ask | “Would you be a reference for me?” | “Would you feel comfortable serving as a strong, positive reference for me?” |
Role Context | “I’m applying for a job at a company.” | “I’m interviewing for a Senior ML Engineer role at AtlasAI, focused on production ML systems and real-time inference.” |
Timeline | “Sometime next week maybe?” | “The hiring team plans to reach out between March 12–15, 2026.” |
Supporting Materials | None attached | Resume PDF, job description link, and 3 bullet points on shared projects |
Closing | “Thanks!” | “Please let me know if you’d like any additional details. I completely understand if timing doesn’t work out.” |
Following Up, Saying Thank You, and Maintaining Relationships
Follow-up and gratitude matter for long-term career growth, especially in the relatively small AI/ML community where relationships often span multiple companies over many years.
Follow-up best practices:
Send a polite follow-up email if you have not heard back after three to four business days. Give an easy out, for example, “I understand if you are too busy or prefer not to serve as a reference. No hard feelings at all.”
Once they agree, send a short thank you note confirming the details and expressing appreciation.
After the interview process concludes, whether you get the offer or not, send another thank you sharing the outcome. If you accept, mention the role, team, and approximate start date.
Keeping relationships strong long-term:
Share major career updates occasionally: a new publication, a notable project shipped, or a promotion.
Don’t only reach out when you need something. A brief “saw your paper and thought it was great” message keeps the connection warm.
Offer to return the favor when colleagues reach out for references or introductions.
How AI Is Changing Hiring and Where References Still Matter
Many AI-focused companies now use AI tools in hiring to screen resumes, prioritize candidates, and schedule interviews. These tools help human resources teams manage high volumes efficiently, but they do not replace human judgment on what makes a candidate truly exceptional.
Despite automation, human references remain a critical signal for validating judgment, teamwork, and ownership, qualities that are hard to infer from a CV or coding task alone. A reference check can reveal how you handled ambiguity on a complex project, how you communicated with cross-functional partners, or how you responded when a production system failed at 2 AM.
For AI engineers and researchers specifically, employers often lean on references to understand real impact and ethics. Did you design safe evaluation protocols for LLMs? How did you handle data privacy concerns? How did you lead the incident response when a model produced unexpected outputs? These nuanced questions require human insight.
See references as a way to bring your full story and character into the process, beyond what an ATS or automated screen might capture. A good reference can speak highly of work that does not fit neatly into resume bullet points.
How Fonzi Uses AI in Hiring With Candidates at the Center
Fonzi is a curated talent marketplace designed specifically for AI engineers, ML researchers, infra engineers, and LLM specialists. Our candidate-first philosophy means we build tools that respect your time and showcase your genuine strengths.
Fonzi uses AI to match candidates and companies more intelligently, based on tech stacks, research interests, compensation targets, and location preferences, while keeping humans in the loop for final decisions. We do not rely on opaque automated scores that leave you guessing why you were not selected.
Our matching process relies on validated experience, portfolio signals, and references to deepen trust between candidates and companies. When a company sees your profile, they understand what you have actually built and shipped, not just keywords on a resume.
Fonzi focuses on reducing bias in the hiring process. Matching is based on skills, contributions, and preferences, not school pedigree or who shouts loudest on social media. Most people find that this approach leads to higher-quality conversations and faster decisions.
Our tools give both candidates and hiring teams more clarity on fit, timelines, and expectations, so reference checks feel purposeful rather than perfunctory.
Match Day on Fonzi: High-Signal Intros Instead of Endless Applications
Match Day simplifies the job search for senior AI talent by flipping the traditional application model. Instead of sending dozens of applications into the void, you complete a detailed profile and receive curated intros from companies that already know they are interested.
Candidates share their skills, projects, location preferences, and salary expectations. On Match Day, companies who match your criteria reach out directly. They have already reviewed high-signal information such as your open-source work, research publications, and infra achievements.
This approach shortens timelines dramatically. Interviews and reference checks often happen in a tight, clearly communicated time frame rather than dragging on for months. Companies are explicit upfront about whether they will need references, so you can line these up using the email templates in this article.
Preparing for Technical Interviews While Managing References
Balancing interview preparation with reference logistics requires intentional planning, especially when you are deep in the interview process for multiple companies.
Block out time early in your job search to refresh practical skills (such as system design for ML infra, LLM prompt engineering), and reading recent papers, and to pre-draft reference emails. Having templates ready means you can send personalized requests within hours of a company asking.
Maintain a simple document listing potential references, their contact details, your professional relationship notes, and the last time you contacted them. When a request comes, you can quickly personalize an email template without starting from scratch.
Prepare 3–5 bullet points about your recent major projects with metrics. These bullet points serve double duty: you will use them in interviews to answer questions about your impact, and you will include them in reference request emails to remind contacts of your shared work.
Conclusion
Structured, respectful reference emails are a career-long skill that pay off in every job search. The AI and ML job market moves quickly, and candidates who prepare references in advance move through hiring more smoothly.
Choosing the right people, asking clearly, and providing context makes it easier for references to advocate strongly for you. A reference request is not a burden; it is an opportunity for someone who knows your work to champion your contributions.
Take action today: adapt a template from this article and reach out to at least one former manager or collaborator. You will be ready to respond confidently when the next hiring team asks for references.
FAQ
How do I write an email asking someone to be my reference?
What should I include in a reference request email to make it easy to say yes?
Is it okay to ask for a reference by email instead of in person?
How do I ask a former boss for a reference if we haven’t spoken in a while?
What do I do if someone doesn’t respond to my reference request email?
