How to Network at Events: 10 Tips for Meaningful Connections
By
Liz Fujiwara
•
Feb 3, 2026
Picture NeurIPS 2026 in San Diego. After sessions end, informal conversations take over. In lobby bars and coffee lines, engineers and researchers connect. By the end of the week, some of these conversations turn into job offers that never appeared on a job board.
For AI and ML engineers, many of the most compelling roles are filled through warm introductions, conference meetups, and curated hiring events. The hidden job market is where hiring often begins.
Networking can feel uncomfortable for technical professionals, but when approached deliberately, it becomes a practical way to access roles, collaborators, and mentors. This article shares 10 practical tips to help you turn conversations into concrete offers faster.
Key Takeaways
Fonzi AI uses bias-audited, fraud-checked, human-in-the-loop AI to shorten hiring cycles and turn conversations into real offers.
Effective networking combines preparation, presence, and timely follow-up, with clear goals, a concise pitch, and outreach within 24 to 48 hours.
The article closes with practical scripts, a comparison table, and FAQs designed for technical candidates who want actionable guidance.
1. Set a Clear Networking Goal Before Every Event
“I’m going to network” is too vague to be useful. Vague goals lead to aimless wandering, random badge collecting, and the exhausted feeling of having talked to many people but connected with no one. Specific goals make events less overwhelming and more effective.
Before your next event, define exactly what you want to accomplish. For AI and ML candidates, clear goals might look like:
Meet three people working on LLM infrastructure at Series A to C startups
Find one hiring manager building a retrieval-augmented generation pipeline
Get feedback on a portfolio project about multimodal models from someone who has shipped similar work
Identify two teams actively hiring remote senior ML engineers
Your goals can fall into different categories:
Learning focused: Understand how teams implement RLHF in production or learn about emerging evaluation frameworks for LLMs
Opportunity focused: Identify two to three companies where your skills align with open roles
Relationship focused: Reconnect with two past colleagues in person or meet one potential mentor in your specialty
Setting goals also helps introverts manage energy. Instead of unstructured mingling, which drains many technical professionals, you have a clear mission. You are not wandering. You are executing a plan. That reframe can make networking feel more manageable and purposeful.
Fonzi’s onboarding follows this approach. Candidates define role preferences, compensation bands, and company stage up front. This clarity helps Match Day conversations start with genuine alignment rather than generic introductions.
2. Craft a Technical Elevator Pitch That Sounds Like You
At networking events, you’ll answer “So what do you do?” dozens of times. Having a crisp, rehearsed answer reduces anxiety and makes a positive impression. A rambling or uncertain response signals that you have not thought clearly about your own work, which is not the message you want to send.
Structure your pitch in three parts:
Label: What you are, your role or title
Focus: What you actually work on, the specific domain or problem
Impact: What you have shipped or improved, with concrete results
Here are three example pitches tailored for AI and ML engineers:
Senior ML Engineer (Recommendation Systems):
“I’m a senior ML engineer specializing in recommendation systems. For the past two years, I’ve been building ranking models at [Company] that power personalized content feeds for 50 million users. Most recently, I led a project that improved click-through rates by 18 percent through better feature engineering and a new embedding approach.”
LLM Infrastructure Engineer:
“I’m an infrastructure engineer focused on LLM deployment and serving. I work on the systems that get large models into production, including latency optimization, cost management, and reliability at scale. At my current company, I reduced inference costs by 40 percent while cutting p99 latency in half.”
Applied Researcher (Multimodal Alignment):
“I’m an applied researcher working on multimodal alignment, getting vision and language models to work together effectively. I published at ICLR last year on cross-modal evaluation and have been exploring how to make these systems more reliable in real-world deployment.”
Adapt your pitch based on your audience. With deeply technical engineers, lean into specifics such as frameworks, architectures, or techniques. With founders or product leaders, emphasize business impact and translate technical wins into outcomes they care about.
This same pitch becomes the backbone of a high-signal Fonzi profile. Hiring managers review profiles before Match Day, so conversations start further along without repeating the basics.
3. Do Your Homework: Research People and Companies in Advance
The most productive networking at events like ICML, specialized AI meetups, or industry conferences is deliberate, not random. Walking in cold and hoping to run into the right people is inefficient. Walking in with a target list and prepared talking points is how meaningful relationships form.
Before the event, take these research steps:
Scan the attendee or speaker list. Most conferences publish this in advance through an event app. Identify five to ten people you want to meet.
Search LinkedIn strategically. Look for titles like “Staff ML Engineer” or “Head of AI” at companies you admire, then see if they are speaking or attending.
Read recent work. Skim one or two blog posts, arXiv papers, or product launches from people on your list to build context.
Use event hashtags and apps. Tags like #ICLR2026 or #LLMDevSummit surface side events, office hours, and sponsor booths where hiring managers spend time.
Prepare one or two tailored questions for each person or company, such as:
“I read your post on scaling retrieval pipelines. How did you handle the cold-start problem?”
“Your team’s approach to model evaluation beyond benchmarks stood out. What is driving that focus?”
“How is your organization thinking about data governance as training datasets scale?”
Showing genuine interest and asking specific questions about someone’s work sets you apart from attendees who collect contacts without context.
4. Start Conversations Without Awkwardness: Practical Openers and Scripts
Even senior engineers can freeze when approaching a group at a crowded mixer or poster session. The fear of interrupting or saying something awkward is real. But everyone is there to connect. Approaching with polite curiosity is expected, not intrusive.
Context-aware openers for 2026 events:
“That talk on RLHF alignment was dense. What was your biggest takeaway?”
“I noticed your badge says [Company]. I’m curious how your team is evaluating LLMs beyond standard benchmarks.”
“I just came from the poster session on multi-agent systems. Did you catch any of those, or did anything stand out?”
“What brought you to this event? Is your team exploring [topic] right now?”
Targeting strategies:
Look for open groups of two to three people at receptions. Gaps in body language often signal they are receptive to someone joining.
Approach people standing alone near coffee stations, food tables, or charging areas. They are often waiting to meet someone.
Avoid interrupting tight circles that are deep in conversation. Wait for a natural pause or look for more accessible groups.
Scripts for different contexts:
After a talk:
“Great talk. I especially liked your point about [specific detail]. I work on something similar and was curious how you handled [specific challenge].”
At a sponsor booth:
“I’ve been following [Company] for a while. What problems are you hiring for right now? Are you building out the inference team?”
Via event app message:
“Hi [Name], I saw you’re speaking on [topic] tomorrow. I work on related problems at [Company or independently]. I’d love to grab coffee and compare notes if you have 15 minutes.”
Networking is much smoother when these scripts are ready. You are not improvising under pressure. You are executing a plan.
5. Listen Like an Engineer: Ask Better Questions, Talk Less
Reframe networking as debugging someone else’s world. Your goal isn’t to give a monologue about yourself. It’s to understand their constraints, use cases, and objectives. The best networkers are strong listeners who ask questions that uncover how they can be helpful.
High-signal questions for AI and ML candidates:
“How does your team move models from research to production? What does the handoff look like?”
“Beyond accuracy metrics, how are you measuring model performance? What matters most to your stakeholders?”
“How is your organization approaching bias and safety in deployed systems?”
“What’s the biggest infrastructure bottleneck you’re dealing with right now?”
“If you could add one ML engineer tomorrow, what problem would you want them to have?”
Weak vs. strong questions:
Weak Question | Stronger Alternative |
“Are you hiring?” | “What problems is your team focused on solving this quarter?” |
“What do you do?” | “What’s the most interesting background project you’re working on right now?” |
“Is AI going to take over?” | “How is your team thinking about the balance between automation and human oversight?” |
Active listening techniques:
Paraphrase what you hear: “So if I understand correctly, your main challenge is…”
Notice nonverbal cues. If someone seems rushed, keep it brief.
Put your phone away. Checking notifications during a conversation breaks rapport.
Ask follow-up questions that build on their answers.
This approach fosters meaningful conversation instead of superficial exchanges. The other person feels heard, and you gain valuable insights.
6. Use Technology Wisely: Business Cards, QR Codes, and Match Day
Networking tools in 2026 combine physical and digital formats. You’ll see printed cards, LinkedIn QR codes, event apps with chat features, and curated marketplaces like Fonzi AI. The savvy networker uses all of them strategically.
The hybrid approach:
Bring business cards, but keep them simple. A clean design with essential information works better than a cluttered card with too much text. Also prepare a scannable QR code on your phone linking to your GitHub, portfolio, or LinkedIn profile.
What to include on cards for engineers:
Name
Role (e.g., “Senior ML Engineer”)
Email
LinkedIn URL
Optional: A short line like “LLM infra & evaluation” or a link to your portfolio
At tech-heavy events, most people prefer exchanging cards digitally or scanning codes. Some business owners and industry veterans still expect physical cards, so having both options covers all scenarios.
Match Day as structured event networking:
Fonzi’s Match Day functions as a “networking event in a box.” Instead of hoping to bump into the right hiring manager at a conference, you apply once, get vetted, and join a scheduled hiring event. Vetted companies are matched with vetted candidates for 48 hours of high-intent conversations and rapid offers.
Companies commit to salary ranges before Match Day, so you are not wasting time on roles that don’t meet your expectations. It is networking with the noise removed.
Comparison: Traditional Networking vs Cold Outreach vs Fonzi Match Day
Here’s how three common approaches to job search networking compare across key dimensions:
Dimension | Traditional In-Person Networking | Cold Outreach (LinkedIn/Discord) | Fonzi AI Match Day |
Typical time to first interview | Weeks to months | Weeks (often no response) | 24–48 hours during event |
Signal vs. noise | Medium; depends on event quality | Low; high volume, low response rates | High; curated matches only |
Candidate preparation required | High; research, travel, attendance | Medium; message crafting | Low; profile once, matched automatically |
Control over salary expectations | None until late in process | None | Full; ranges disclosed upfront |
Bias & fairness safeguards | None | None | Bias-audited evaluations, fraud detection |
Emotional load for introverts | High; extended social exposure | Medium; asynchronous but rejection-heavy | Low; structured, time-boxed interactions |
7. Manage Energy and Anxiety: Networking for Introverts and Deep-Work People
Many high-performing AI and ML engineers are introverted or prefer long stretches of focused work over continuous small talk. Multi-day conferences can feel socially exhausting. The key is strategic pacing rather than trying to follow an extrovert’s playbook.
Practical pacing strategies:
Set a per-hour connection target. Aim for two to three quality chats instead of 15 card swaps
Schedule quiet breaks. Block 30 minutes between sessions to recharge alone
Avoid packing every time slot. Leave buffer for hallway conversations, which are often more valuable than formal sessions
Use the event app to pre-schedule 1:1 meetings to reduce cold-approaching
Tips for managing social anxiety:
Prepare 3–4 default questions you can deploy when your mind goes blank
Attend smaller side events; workshops, birds-of-a-feather meetups, or niche dinners have lower pressure than massive receptions
Stand near food or coffee stations where starting a meaningful conversation happens naturally
Arrive early when the room is less crowded and it’s easier to meet people one-on-one
Anchor events reduce pressure:
Poster sessions, workshops, and topic-specific roundtables provide instant common ground. You can discuss the poster in front of you or the workshop problem you just solved, avoiding the need for manufactured conversation starters. These formats favor depth over breadth, ideal for engineers who find small talk exhausting.
8. Exit Conversations Gracefully and Follow Up Fast
Knowing how to leave conversations politely lets you meet more people without burning bridges. Lingering too long wastes time, and exiting abruptly feels rude. The goal is a graceful transition that leaves a positive impression.
The exit formula:
Brief thanks for the conversation
Reference something specific you enjoyed or learned
Clear next step (swap contact info, promise to send a resource, suggest connecting on LinkedIn)
Example exit lines:
“This was really helpful. I loved hearing about your approach to evaluation. I don’t want to monopolize your time, but let’s connect on LinkedIn. I’ll send over that paper I mentioned.”
“Great chatting with you. I’m going to grab some coffee and catch the next session, but I’d love to continue this. Can I get your card?”
“I should let you meet other people, but this was one of the best conversations I’ve had today. I’ll follow up with that repo link I mentioned.”
The follow-up timeline:
Aim for 24–48 hours after the event while the conversation is fresh. Your message should:
Reference something specific you discussed
Include one small, relevant value-add, such as a resource, article, or introduction
Keep it short, two to three sentences is enough
Example follow-up message:
“Hi [Name], it was great meeting you at [Event] yesterday. I really enjoyed our discussion about scaling inference for LLMs. Your point about batching strategies stuck with me. Here’s that blog post I mentioned on dynamic batching. I would love to stay in touch as you build out the team.”
9. Turn One-Off Chats into Long-Term Career Capital
The real impact of networking often appears months or years later. A brief connection at a poster session can turn into a research collaboration. A hiring manager without an open role in 2026 may reach out in 2027 with a founding engineer opportunity. Weak ties grow stronger through consistent, low-effort maintenance.
Simple maintenance habits:
When someone publishes a paper, ships a product, or raises a funding round, send a short congrats message
Like or comment thoughtfully on their LinkedIn posts; don’t just scroll past
Share relevant articles or research that connects to past conversations
Check in occasionally without asking for anything
Build a lightweight network log:
After big events, create a simple doc or note listing:
Who you met
What you talked about
Their interesting background details
Potential next steps or reasons to reconnect
This takes 10 minutes and prevents the “who was that person again?” problem six months later.
Become a connector:
One of the most valuable networking skills is introducing people who could help each other. For example, notice a data engineer struggling with pipeline issues at dinner? Connect them with the startup founder you met who needs that expertise. Helping others build connections creates social capital without expecting an immediate return.
10. Use AI in Your Favor: How Fonzi Makes Networking More Fair and Efficient
AI is already common in hiring, from resume screening and assessment tools to coding challenges and interview scheduling. For candidates, this can feel opaque or unfair. You submit an application and it disappears into an algorithmic black box, leaving you to wonder if a human ever saw your work.
Fonzi AI takes a different stance. The platform uses AI to reduce noise, not replace humans:
Fraud detection keeps out fake profiles and ensures you’re competing against real candidates
Bias-audited evaluations help create a more level playing field
Automated scheduling eliminates the back-and-forth that slows down hiring
Human-led decisions mean actual hiring managers and CTOs make the final calls
The Match Day flow from a candidate’s perspective:
Apply once and create your profile
Get vetted by Fonzi’s team
Join a scheduled Match Day event
See upfront salary ranges from participating companies
Talk directly with hiring managers and CTOs over 48 hours
Receive offers and make decisions
Salary transparency is built into the community. Companies commit to compensation bands before Match Day starts, so you can focus on opportunities that align with your expectations. No more discovering in round three that a role pays 40 percent less than you need.
The stack approach:
Events, personal networking, and Fonzi work together:
Learn and connect at conferences and meetups across various industries
Build relationships through the maintenance habits above
Convert momentum through Fonzi’s structured marketplace when you’re ready to move
This combination provides the best of both worlds: the serendipity of in-person connections and the efficiency of a curated hiring process.
Conclusion
Effective networking for AI, ML, and infra engineers is about intention, preparation, and consistent follow-through, not being the loudest person in the room. Focus on clear goals, a sharp elevator pitch, research, thoughtful questions, and follow-up within 24 to 48 hours.
Real opportunities come from meaningful conversations where you show your thinking, understand the other person’s challenges, and maintain the relationship over time. It’s the few strong connections you nurture that create career capital.
Turn your networking momentum into real offers. Create a free candidate profile on Fonzi AI, get vetted, and join the next Match Day. For more guidance on AI hiring, technical interviews, and building a standout engineering career, explore the Fonzi blog.
