Mark Gurman reported today that Apple will finally unveil the new Siri this June at its annual WWDC conference.

The post Apple to Unveil the New Siri at WWDC 2026 in June appeared first on Thurrott.com.

Why developers need a fast lane from research → prototypes

AI engineering has a speed problem, but it is not a shortage of announcements. The hard part is turning research into a useful prototype before the next wave of models, tools, or agent patterns shows up.

That gap matters. AI engineers want to compare quality, latency, and cost before they wire a model into a product. Full-stack teams want to test whether an agent workflow is real or just demo. Platform and operations teams want to know when an experiment can graduate into something observable and supportable.

Microsoft makes that case directly in introducing Microsoft Foundry Labs: breakthroughs are arriving faster, and time from research to product has compressed from years to months.

If you build real systems, the question is not "What is the coolest demo?" It is "Which experiments are worth my next hour, and how do I evaluate them without creating demo-ware?" That is where Microsoft Foundry Labs becomes interesting.

What is Microsoft Foundry Labs?

Microsoft Foundry Labs is a place to explore early-stage experiments and prototypes from Microsoft, with an explicit focus on research-driven innovation. The homepage describes it as a way to get a glimpse of potential future directions for AI through experimental technologies from Microsoft Research and more. The announcement adds the operating idea: Labs is a single access point for developers to experiment with new models from Microsoft, explore frameworks, and share feedback.

That framing matters. Labs is not just a gallery of flashy ideas. It is a developer-facing exploration surface for projects that are still close to research: models, agent systems, UX ideas, and tool experiments. Here's some things you can do on Labs: 

1. Play with tomorrow’s AI, today: 30+ experimental projects—from models to agents—are openly available to fork and build upon, alongside direct access to breakthrough research from Microsoft.

2. Go from prototype to production, fast: Seamless integration with Microsoft Foundry gives you access to 11,000+ models with built-in compute, safety, observability, and governance—so you can move from local experimentation to full-scale production without complex containerization or switching platforms.

3. Build with the people shaping the future of AI: Join a thriving community of 25,000+ developers across Discord and GitHub with direct access to Microsoft researchers and engineers to share feedback and help shape the most promising technologies.

What Labs is not: it is not a promise that every project has a production deployment path today, a long-term support commitment, or a hardened enterprise operating model.

Spotlight: a few Labs experiments worth a developer's attention

• Phi-4-Reasoning-Vision-15B: A compact open-weight multimodal reasoning model that is interesting if you care about the quality-versus-efficiency tradeoff in smaller reasoning systems.
• BitNet: A native 1-bit large language model that is compelling for engineers who care about memory, compute, and energy efficiency.
• Fara-7B: An ultra-compact agentic small language model designed for computer use, which makes it relevant for builders exploring UI automation and on-device agents.
• OmniParser V2: A screen parsing module that turns interfaces into actionable elements, directly relevant to computer-use and UI-interaction agents.

If you want to inspect actual code, the Labs project pages also expose official repository links for some of these experiments, including OmniParser, Magentic-UI, and BitNet.

Labs vs. Foundry: how to think about the boundary

The simplest mental model is this: Labs is the exploration edge; Foundry is the platform layer.

The Microsoft Foundry documentation describes the broader platform as "the AI app and agent factory" to build, optimize, and govern AI apps and agents at scale. That is a different promise from Labs. Foundry is where you move from curiosity to implementation: model access, agent services, SDKs, observability, evaluation, monitoring, and governance.

Labs helps you explore what might matter next. Foundry helps you build, optimize, and govern what matters now. Labs is where you test a research-shaped idea. Foundry is where you decide whether that idea can survive integration, evaluation, tracing, cost controls, and production scrutiny.

That also means Labs is not a replacement for the broader Foundry workflow. If an experiment catches your attention, the next question is not "Can I ship this tomorrow?" It is "What is the integration path, and how will I measure whether it deserves promotion?"

What's real today vs. what's experimental

• Real today: Labs is live as an official exploration hub, and Foundry is the broader platform for building, evaluating, monitoring, and governing AI apps and agents.
• Experimental by design: Labs projects are presented as experiments and prototypes, so they still need validation for your use case.

A developer's lens: Models, Agents, Observability

What makes Labs useful is not that it shows new things. It is that it gives developers a way to inspect those things through the same three concerns that matter in every serious AI system: model choice, agent design, and observability.

Diagram description: imagine a loop with three boxes in a row: Models, Agents, and Observability. A forward arrow runs across the row, and a feedback arrow loops from Observability back to Models. The point is that evaluation data should change both model choices and agent design, instead of arriving too late.

Models: what to look for in Labs experiments

If you are model-curious, Labs should trigger an evaluation mindset, not a fandom mindset. When you see something like Phi-4-Reasoning-Vision-15B or BitNet on the Labs homepage, ask three things: what capability is being demonstrated, what constraints are obvious, and what the integration path would look like.

This is where the Microsoft Foundry Playgrounds mindset is useful even if you started in Labs. The documentation emphasizes model comparison, prompt iteration, parameter tuning, tools, safety guardrails, and code export. It also pushes the right pre-production questions: price-to-performance, latency, tool integration, and code readiness.

That is how I would use Labs for models: not to choose winners, but to generate hypotheses worth testing. If a Labs experiment looks promising, move quickly into a small evaluation matrix around capability, latency, cost, and integration friction.

Agents: what Labs unlocks for agent builders

Labs is especially interesting for agent builders because many of the projects point toward orchestration and tool-use patterns that matter in practice. The official announcement highlights projects across models and agentic frameworks, including Magentic-One and OmniParser v2. On the homepage, projects such as Fara-7B, OmniParser V2, TypeAgent, and Magentic-UI point in a similar direction: agents get more useful when they can reason over tools, interfaces, plans, and human feedback loops.

For working developers, that means Labs can act as a scouting surface for agent patterns rather than just agent demos. Look for UI or computer-use style agents when your system needs to act through an interface rather than an API. Look for planning or tool-selection patterns when orchestration matters more than raw model quality.

My suggestion: when a Labs project looks relevant to agent work, do not ask "Can I copy this architecture?" Ask "Which agent pattern is being explored here, and under what constraints would it be useful in my system?"

Observability: how to experiment responsibly and measure what matters

Observability is where prototypes usually go to die, because teams postpone it until after they have something flashy. That is backwards. If you care about real systems, tracing, evaluation, monitoring, and governance should start during prototyping.

The Microsoft Foundry documentation already puts that operating model in plain view through guidance for tracing applications, evaluating agentic workflows, and monitoring generative AI apps. The Microsoft Foundry Playgrounds page is also explicit that the agents playground supports tracing and evaluation through AgentOps.

At the governance layer, the AI gateway in Azure API Management documentation reinforces why this matters beyond demos. It covers monitoring and logging AI interactions, tracking token metrics, logging prompts and completions, managing quotas, applying safety policies, and governing models, agents, and tools. You do not need every one of those controls on day one, but you do need the habit: if a prototype cannot tell you what it did, why it failed, and what it cost, it is not ready to influence a roadmap.

"Pick one and try it": a 20-minute hands-on path

Keep this lightweight and tool-agnostic. The point is not to memorize a product UI. The point is to run a disciplined experiment.

1. Browse Labs and pick an experiment aligned to your work. Start at Microsoft Foundry Labs and choose one project that is adjacent to a real problem you have: model efficiency, multimodal reasoning, UI agents, debugging workflows, or human-in-the-loop design.
2. Read the project page and jump to the repo or paper if available. Use the Labs entry to understand the claim being made. Then read the supporting material, not just the summary sentence.
3. Define one small test task and explicit success criteria. Keep it concrete: latency budget, accuracy target, cost ceiling, acceptable safety behavior, or failure rate under a narrow scenario.
4. Capture telemetry from the start. At minimum, keep prompts or inputs, outputs, intermediate decisions, and failures. If the experiment involves tools or agents, include tool choices and obvious reasons for failure or recovery.
5. Make a hard call. Decide whether to keep exploring or wait for a stronger production-grade path. "Interesting" is not the same as "ready for integration."

Minimal experiment logger (my suggestion): if you want a lightweight way to avoid demo-ware, even a local JSONL log is enough to capture prompts, outputs, decisions, failures, and latency while you compare ideas from Labs.

import json
import time
from pathlib import Path

LOG_PATH = Path("experiment-log.jsonl")


def record_event(name, payload):
	# Append one event per line so runs are easy to diff and analyze later.
	with LOG_PATH.open("a", encoding="utf-8") as handle:
		handle.write(json.dumps({"event": name, **payload}) + "\n")


def run_experiment(user_input):
	started = time.time()
	try:
		# Replace this stub with your real model or agent call.
		output = user_input.upper()
		decision = "keep exploring" if len(output) < 80 else "wait"
		record_event(
			"experiment_result",
			{
				"input": user_input,
				"output": output,
				"decision": decision,
				"latency_ms": round((time.time() - started) * 1000, 2),
				"failure": None,
			},
		)
	except Exception as error:
		record_event(
			"experiment_result",
			{
				"input": user_input,
				"output": None,
				"decision": "failed",
				"latency_ms": round((time.time() - started) * 1000, 2),
				"failure": str(error),
			},
		)
		raise


if __name__ == "__main__":
	run_experiment("Summarize the constraints of this Labs project.")

That script is intentionally boring. That is the point. It gives you a repeatable, runnable starting point for comparing experiments without pretending you already have a full observability stack.

Practical tips: how I evaluate Labs experiments before betting a roadmap on them

• Separate the idea from the implementation path. A strong research direction can still have a weak near-term integration story.
• Test one workload, not ten. Pick a narrow task that resembles your production reality and see whether the experiment moves the needle.
• Track cost and latency as first-class metrics. A novel capability that breaks your budget or response-time envelope is still a failed fit.
• Treat agent demos skeptically unless you can inspect behavior. Tool calls, traces, failure cases, and recovery paths matter more than polished output.

Common pitfalls are predictable here.

• Do not confuse a research win with a deployment path. Labs is for exploration, so you still need to validate integration, safety, and operations.
• Do not evaluate with vague prompts. Use a narrow task and explicit success criteria, or you will end up comparing vibes instead of outcomes.
• Do not skip telemetry because the prototype is small. If you cannot inspect failures early, the prototype will teach you very little.
• Do not ignore known limitations. For example, the Fara-7B project page explicitly notes challenges on more complex tasks, instruction-following mistakes, and hallucinations, which is exactly the kind of constraint you should carry into evaluation.

What to explore next

Azure AI Foundry Labs matters because it gives developers a practical way to explore research-shaped ideas before they harden into mainstream patterns. The smart move is to use Labs as an input into better platform decisions: explore in Labs, validate with the discipline encouraged by Foundry playgrounds, and then bring the learnings back into the broader Foundry workflow.

• Takeaway 1: Labs is an exploration surface for early-stage, research-driven experiments and prototypes, not a blanket promise of production readiness.
• Takeaway 2: The right workflow is Labs for discovery, then Microsoft Foundry for implementation, optimization, evaluation, monitoring, and governance.
• Takeaway 3: Tracing, evaluations, and telemetry should start during prototyping, because that is how you avoid confusing a compelling demo with a viable system.

If you are curious, start with Microsoft Foundry Labs, read the official context in Introducing Microsoft Foundry Labs, and then map what you learn into the platform guidance in Microsoft Foundry documentation.

Try this next

• Open Microsoft Foundry Labs and choose one experiment that matches a real workload you care about.
• Use the mindset from Microsoft Foundry Playgrounds to define a small validation task around quality, latency, cost, and safety.
• Write down the minimum telemetry you need before continuing: inputs, outputs, decisions, failures, and token or cost signals.
• Read the relevant operating guidance in AI gateway in Azure API Management if your experiment may eventually need monitoring, quotas, safety policies, or governance.
• Promote only the experiments that can explain their value clearly in a Foundry-shaped build, evaluation, and observability workflow.

Register now and save your spot for the chance to ask questions and share feedback with hundreds of Microsoft executives, engineers, and product leaders. By making your voice heard, your feedback will help shape the technologies that you and countless others use every day.

Plus, you’ll hear from icons of Intelligent Work during inspiring keynotes, sessions, and AMAs, including:  

• Jeff Teper, President, Microsoft 365 Collaborative Apps & Platforms, Microsoft  

• Ryan Cunningham, Corporate Vice President, Power Platform 

• Vasu Jakkal, Corporate Vice President, Microsoft Security 

• Rohan Kumar, Corporate Vice President, Microsoft Security, Purview & Trust 

• Dr. Jaime Teevan, Chief Scientist & Technical Fellow, Microsoft 

View full speaker list > 

What: Microsoft 365 Community Conference 2025  

• Register today | Note: Use the SAVE150 discount code to save $150 USD.  

• Content snapshot: 5 Microsoft keynotes + 1 Ask Microsoft Anything (AMA) | 200+ overall sessions  Microsoft-led sessions | 21 full-day workshops (pre-day and post)  

• Microsoft is sending over 150+ product makers to present and engage.  

• Review all sessions, workshops, and speakers.  

When & where: April 21-23, 2026 

• In-person: Loews Sapphire Falls and Loews Royal Pacific Resorts in Orlando, Florida 

• Pre-day workshops on Sunday, April 19 and Monday, April 20. 

Details:

• Twitter & hashtag: @M365CONF | #M365Con  

• Follow us on LinkedIn: M365Community Conference LinkedIn Company Page  

• Cost: $1,850- full conference (Includes 3 continental breakfasts, 3 lunches, a T-shirt, and backpack (additional costs for full-day workshops). Please visit the M365 Community Conference for full pricing.   

• Microsoft Customer Discount:   

• Use CODE: SAVE150 for $150 USD off registration. 

Why attend?   

• Stay Ahead of Innovation: Gain exclusive insights into the latest advancements in AI and Microsoft’s Product Roadmap. Empowering you to drive transformation within your organization and stay at the forefront of technology.  

• Hands-On Learning: Participate in practical sessions and workshops that provide actionable strategies and skills to optimize workflows, enhance productivity, and achieve business success.  

• Connect with Experts: Build invaluable connections with Microsoft product makers, MVPs, industry leaders, and a global community of like-minded professionals who share your passion for technology.  

• Transform Your Impact: Leave with a deeper understanding of Microsoft tools, clear strategies for leveraging AI in your work, and a roadmap for achieving your goals and enhancing your team's success.  

• Build Community: Build community by connecting with others, sharing experiences, and forming meaningful relationships.  

• Check out Jeff Teper's awesome statement on The ultimate Microsoft 365 community event is back| Microsoft 365 Blog.  

• FUN: Join us in our community lounge for networking, hanging out, meet and greets with Microsoft and Community All-Stars, surprise and delight treats, stickers and raffles to win grand prizes! 

Are you in?

Register today! Don't miss your chance to join the largest M365 focused conference of the year.  

Ways to save: 

1. Use CODE: SAVE150 for $150 USD off registration.
2. M365CON26 Grant Program (For Non-profits & Small Business)

See if you qualify for a deep discount or free pass for Microsoft 365 Community Conference. Small Businesses & Non-profits are eligible to apply for a free pass grant at www.change3e.com/grants (government organizations do not quality).

Read on for more details about what to expect, and we hope to see you there! 

High-level conference schedule

*subject to change – check back for updates 

Sunday, April 19, 2026 – Workshop Day  

• 9:00 AM- 4:00 PM Workshops  

Monday, April 20, 2026 – Workshop Day  

• 9:00 AM- 4:00 PM Workshops  

Tuesday, April 21, 2026 –Conference Session Day  

• 7:00 AM- 8:15 AM Continental Breakfast  

• 8:30 AM- 5:00 PM Keynotes and Sessions  

• 5:00 PM- 6:30 PM Conference Opening Recep7on  

Wednesday, April 22, 2026 – Conference Session Day  

• 7:00 AM- 8:00 AM Continental Breakfast  

• 8:00 AM - 5:00 PM Keynotes and Sessions  

• 5:00 PM - 6:15 PM Networking and T-shirt pick up in Expo Hall  

• 8:00 PM - 10:00 PM Attendee Party Universal Islands of Adventure 

Thursday April 23, 2026 – Conference Session Day  

• 7:30 AM– 8:45 AM Continental Breakfast  

• 9:00 AM – 4:30 PM Keynotes and Sessions Friday 

April 24, 2026 – Workshop Day  

• 8:00 AM - 3:00 PM Workshops 

Check out the keynotes and speakers 

Catch the speakers and sessions that matter most to you, and your customers.   

• Keynote: Building for the Future: Microsoft 365, Agents and AI, What's New and What's Next
  Jeff Teper, President, Collaborative Apps and Platforms, Microsoft
  Tuesday, April 21, 8:30-9:30 AM 
  
  

• Keynote: Business Apps and Agents
  Ryan Cunningham, Corporate Vice President, Microsoft
  Tuesday, April 21, 9:30-10:20 AM 
  
  

• Keynote: Securing AI: Building Trust in the Era of AI
  Vasu Jakkal, Corporate Vice President, Microsoft Security Business
  Rohan Kumar, Corporate Vice President, Security Purview and Trust, Microsoft
  Wednesday, April 22, 8:00-9:15 AM 
  
  

• Keynote: The Future of Work Fireside Keynote 
  Jamie Teevan, Ph.D., Chief Scientist and Technical Fellow, Microsoft
  Thursday, April 23, 2:15-3:15 PM 

• Keynote: From Momentum to Movement: Where Community Goes Next
  Karuana Gatimu, Director of Customer Advocacy, AI and Collaboration, Microsoft
  Heather Cook, Principal PM Manager, Global Community Evangelism and Events Customer Advocacy Group, Microsoft 365
  Thursday, April 23, 3:15-4:15 PM 

Full session list 

200+ sessions and workshops 

If you’re interested in learning how Microsoft is shaping the future of work—there’s a session for you. Join Microsoft experts, MVPs, and community leaders as you find out about the latest product innovations and AI-powered tools. Discover what’s new with Microsoft 365 Copilot, Teams, SharePoint, OneDrive, Copilot Studio, and much more. Check out the full catalog >

Explore sessions by day: 

Tuesday, April 21, 2026 – Conference Session Day  

Wednesday, April 22, 2026 – Conference Session Day  

Thursday April 23, 2026 – Conference Session Day  

Friday, April 24, 2026 – Workshop Day  

Level up your AI skills and shape the future of your business with Featured Sessions 

Develop the skills needed to thrive as a Frontier Firm through sessions led by innovators, builders, designers, Microsoft product teams, and MVPs shaping collaboration.

• Featured Session: Engage everywhere: communities, events, and storylines in Teams, powered by AI with Murali Sitaram, Steve Nguyen, Jason Mayans, and Jeanette Vikbacka Castaing 

• Featured Session: Your Guide What’s New in Teams: Collaboration, Communication, and Copilot with Ilya Bukshteyn and Chandra Chivukula 

• Planner, Copilot and Agents: AI Assisted Projects with WorkIQ with Howard Crow and Robyn Guarriello 

Lightning Talks: Lightning Talks are fast‑paced, 15‑minute sessions happening live at the Microsoft Booth that are designed to deliver practical insights, real‑world tips, and quick wins you can take back to your team right away. Stop by throughout the day to learn something new, get inspired, and catch a few surprises along the way. You can check the event agenda for session details.

Onsite experiences

Demo Stations: Visit the demo stations in the Microsoft Booth for guided, live walkthroughs led by Microsoft experts who can answer your questions in real time. Each station is designed to support scenario‑based conversations and hands‑on exploration, so you can dive deeper into the tools and solutions that matter most to your organization. The Microsoft Booth will be open anytime the expo hall is open. 

Mini LEGO figurines: Build a mini-LEGO figurine as a takeaway after walking through a demo scenario. It is a quick hands-on moment that adds some play and personality to your conference experience and lets you take some fun swag home. 

Celebrate 25 years of SharePoint 

Tuesday, April 21, 12:30PM – 1:30PM  
Celebrate the SharePoint community with the premiere of a new documentary short film at the conference, complete with a true movie‑premiere experience. Designed to honor the impact of SharePoint and Microsoft 365, this special event brings the community together to capture and celebrate the moment in person.

Tuesday, April 21, 5PM – 6:30PM 
Microsoft Booth
Join a live panel celebrating 25 years of SharePoint and what is next for the platform and community. Expect reflections, stories, and forward-looking conversations tied to the broader SharePoint at 25 celebration.  
 

Surface + Global Skilling Bar: Get hands on with Surface devices while exploring skilling resources designed to help you level up. This activation is being built to connect Surface experiences with AI skilling guidance, including the AI Skills Navigator. Attendees can experience this hands-on learning in the Microsoft booth during expo hall hours. 

Surface giveaways from raffle: Enter the raffle for a chance to win Surface giveaways and other prizes during the event. Winners are selected as part of the community giveaway experience, so check posted instructions onsite and make sure you enter. Each raffle will take place at the close of expo hall hours at the Microsoft Booth 
 

Bring your spirit wear game! 

We’re bringing the fun to the show floor with themed Costume Spirit Days all week long! Come by the booth in your best look, show off your costume, and score an extra raffle ticket for our daily prize drawing. 

Product Roundtable Discussions 

Microsoft wants to hear from you! 

The Microsoft User Experience Research team is hosting a series of interactive sessions focused on AI-powered experiences across Microsoft Teams, OneDrive, SharePoint, Planner, and Viva Engage. We invite you to join us to share your perspectives, explore emerging ideas, and influence what’s next. 

Customer feedback is at the heart of our innovation process. Your participation helps ensure that the experiences we are building are useful, intuitive, and grounded in real organizational needs. 

“At Microsoft, learning from our customers is a privilege and a vital part of our customer-driven innovation process. By sharing your perspectives, you enable us to create solutions that truly meet your needs and enhance your overall experience with Microsoft products.”  

– Marcella Silva, Partner Director of User Experience Research 

What you can expect: 

• Interactive Sessions: Participate in hands-on activities and guided discussions with the Microsoft User Experience Research and Product teams. 

• Expert Insights: Learn directly from the teams shaping the experiences you use every day. 

• Exclusive Previews: Get early visibility into current explorations and future directions.  

• Meaningful Connections: Engage with peers and Microsoft teams in a collaborative environment. 

How to sign up: 

Stay tuned! The sign-up form and detailed session descriptions will be available soon. 

Don’t miss this opportunity to make a difference in the future of Microsoft 365! 

Community meetups  

Community meetups are informal, connection‑first gatherings where attendees can meet others who share a role, interest, or identity. Whether you're looking to exchange ideas, grow your network, or find your people within the Microsoft 365 community, these sessions create space for meaningful conversations and authentic connection beyond the conference sessions.

Women in Tech and Allies Lunch Discussion 

Grab your lunch and network with peers, discuss progress, and focus on how we can continue to make the Microsoft ecosystem the best in the world for supporting women. All individuals are welcome regardless of gender for an open and compassionate conversation on allyship and inclusiveness.  

The discussion will be hosted by Heather Cook, Principal Customer Experience PM and distinguished panelists including: 

• Vasu Jakkal, Corporate Vice President, Microsoft Security Business 

• Jaime Teevan, Chief Scientist & Technical Fellow 

• Lan Ye, Corporate Vice President, Microsoft Teams 

• Sumi Singh, Corporate Vice President, Microsoft Teams 

• Karuana Gatimu, Director, Customer Advocacy - AI & Collaboration, Microsoft 

• Kate Faaland, Program Manager of Data and AI, AvePoint 

Get the MVP experience—and don’t forget the photo op! 

Microsoft MVPs will be on hand throughout the event to share their real-world experience and practitioner insight directly to you. During sessions, our MVPs will share thoughts and brainstorm solutions to challenges as they share practical, scenario‑driven guidance drawn from hands‑on work with Microsoft 365.  

And of course, the program isn’t over without the annual MVP photo!

Register Today! 

You’ll find all this and more at Microsoft 365 Community Conference, April 21–23 in Orlando.  

Register today and don't miss your chance for the largest M365 focused conference of the year. and don't miss your chance for the largest M365 focused conference of the year. 

Ways to save:

1. Use CODE: SAVE150 for $150 USD off registration.
2. M365CON26 Grant Program (For Non-profits & Small Business) 

See if you qualify for a deep discount or free pass for Microsoft 365 Community Conference. Small Businesses & Non-profits are eligible to apply for a free pass grant at www.change3e.com/grants (government organizations do not quality).

Learn more   

Visit M365Conf.com and follow the action on: X/Twitter: @M365Conf / #M365Con, @Microsoft365, @MSFTCopilot, @SharePoint, @OneDrive, @MicrosoftTeams, @MSPowerPlat, @Microsoft365Dev, and @MSFTAdoption. And on our Microsoft Community LinkedIn and Learning YouTube channel.

Want more information? Check out our library of blogs on topics, speakers, community activities and more: 

• The ultimate Microsoft 365 community event returns with Jeff Teper 

• 5 reasons to attend this year’s event 

• What’s new in Microsoft Teams 

• What to expect from the Copilot & AI sessions 

• MVP experiences 

• Get your company engagement up with Viva Engage 

• Women in Tech at M365 Community Conference  

• Level up your data security 

• Keynotes highlight 
  

We hope you will join us in Orlando for an unforgettable week of innovation, inspiration, and community! Get ready to dive into the action with passionate MVPs, Microsoft product leaders, and an incredible lineup of tools like Microsoft 365 Copilot, Teams, SharePoint, OneDrive, Copilot Studio, Power Platform, Planner, and so much more. Let us make memories and shape the future together!  

At the Microsoft 365 Community Conference this April 21–23 in Orlando, we’re proud to spotlight a range of experiences created to support, connect, and celebrate women across the tech community—from leadership panels and networking opportunities to professional development sessions designed to help you grow your impact and your career. Join us for the Women in Tech & Allies Lunch on Wednesday, April 22 sponsored by AvePoint and explore Diversity, Equity, Inclusion, and Professional Development sessions throughout the week.

Be sure to register for the conference so you can dive into all of these activations onsite!

A big thank you to the women on the Microsoft 365 Community Conference Community Content Committee, Susan Hanley, Martina Grom who worked on the community session selection for the show.

Women in Tech & Allies Lunch

Calling all women in tech and allies! Join us on Wednesday April 22, for the Women in Tech & Allies Lunch, your opportunity to network with peers, discuss the unique challenges facing women in the field and brainstorm solutions, and celebrate the talent, expertise, and dexterity that is uniquely you.

This community experience, held at the Microsoft 365 Community Conference, April 21–23 in Orlando, is among the most popular moments of the conference year over year, and we want you to be a part of it. This year, we’re splitting the lunch into two segments: we’ll begin with a dynamic executive panel, which will then lead into a networking activation designed to spark connection, allyship, and actionable insights.

The executive panel will be hosted by Heather Cook, Microsoft’s Principal Customer Experience Manager. She'll be joined by a powerhouse lineup of panelists, including:

• Vasu Jakkal, Corporate Vice President, Microsoft Security Business
• Dr. Jaime Teevan, Chief Scientist & Technical Fellow
• Lan Ye, Corporate Vice President, Microsoft Teams
• Sumi Singh, Corporate Vice President, Microsoft Teams
• Karuana Gatimu, Director, Customer Advocacy - AI & Collaboration, Microsoft
• Kate Faaland, Program Manager of Data and AI, AvePoint

They will discuss how women are utilizing Microsoft’s solutions to shape the Frontier Firm. Heather and the panelists will discuss the challenges they’ve faced as women in tech, as well as support from allies and how innovations in the field have built a foundation for their growth and the growth of their teams. Wristbands for the Women in Tech & Allies Lunch will need to be picked up on-site.

And of course, there will be ample opportunities for fun, networking, and swag! Attendees will walk away with Women in Microsoft 365 enamel pins, as well as Women in Power stickers and some great raffle prizes. Plus, they’ll have the opportunity to network with other attendees over a good meal.

Don't miss out on the Diversity, Equity, Inclusion, & Professional Development sessions

• Fireside Chat: Being an Effective Leader with AI with Karuana Gatimu Director, Customer Advocacy - AI & Collaboration, Microsoft
• Empowering Community Builders: Join MGCI & CommunityDays.org with Heather Cook, Principal PM Manager, Global Community Evangelism & Events Customer Advocacy Group, Microsoft 365, Microsoft, Emily Hove, Global Director, Microsoft Reactor, Microsoft, Bryan Hart, Senior Customer Experience PM, Microsoft
• Leading Workforce Transformation: The Art and Science of Skilling Your People with Karuana Gatimu, Director, Customer Advocacy - AI & Collaboration, Microsoft, Jessie Hwang, Customer Experience PM, Microsoft 365 Customer Advocacy Group, AI & Collaboration, Microsoft
• How To Thrive In Your Tech Career: Managing Burnout and Boredom with Adam Harmetz, Vice President of Product, Microsoft
• Cultivating Trust and Leadership Excellence: Strategies for Respect and Empathy in the Workplace with Heather Cook, Principal PM Manager, Global Community Evangelism & Events Customer Advocacy Group, Microsoft 365, Microsoft
• Be sure to check out all DEI and Professional Development content in our session catalog and in Whova for additional programming to come

Insights, solutions, connection, and good times, what more would you want from an afternoon? Register now and be a part of something special April 21–23. Use the SAVE150 code for $150 USD off current registration pricing.

Privacy and user control remain at the heart of the Android experience. Just as the photo picker made media sharing secure and easy to implement, we are now bringing that same level of privacy, simplicity, and great user experience to contact selection.

A New Standard for Contact Privacy

How It Works

• Granular Data Requests: Apps can specify exactly which fields they need, such as phone numbers or email addresses, rather than receiving the entire contact record.

• Multi-Selection Support: The picker supports both single and multiple contact selections, giving developers more flexibility for features like group invitations.

• Selection Limits: Developers can set custom limits on the number of contacts a user can select at one time.

• Temporary Access: Upon selection, the system returns a Session URI that provides temporary read access to the requested data, ensuring that access does not persist longer than necessary.

• Access to other profiles: When using this new intent, the interface will allow users to select contents from other user profiles such as a work profile, cloned profile or a private space.

// State to hold the list of selected contacts
var contacts by remember { mutableStateOf<List<Contact>>(emptyList()) }

// Launcher for the Contact Picker intent
val pickContact = rememberLauncherForActivityResult(StartActivityForResult()) {
    if (it.resultCode == Activity.RESULT_OK) {
        val resultUri = it.data?.data ?: return@rememberLauncherForActivityResult

        // Process the result URI in a background thread
        coroutine.launch {
            contacts = processContactPickerResultUri(resultUri, context)
        }
    }
}

// Define the specific contact data fields you need
val requestedFields = arrayListOf(
    Email.CONTENT_ITEM_TYPE,
    Phone.CONTENT_ITEM_TYPE,
)

// Set up the intent for the Contact Picker
val pickContactIntent = Intent(ACTION_PICK_CONTACTS).apply {
    putExtra(EXTRA_PICK_CONTACTS_SELECTION_LIMIT, 5)
    putStringArrayListExtra(
        EXTRA_PICK_CONTACTS_REQUESTED_DATA_FIELDS,
        requestedFields
    )
    putExtra(EXTRA_PICK_CONTACTS_MATCH_ALL_DATA_FIELDS, false)
}

// Launch the picker
pickContact.launch(pickContactIntent)

// Data class representing a parsed Contact with selected details
data class Contact(val id: String, val name: String, val email: String?, val phone: String?)

// Helper function to query the content resolver with the URI returned by the Contact Picker.
// Parses the cursor to extract contact details such as name, email, and phone number
private suspend fun processContactPickerResultUri(
    sessionUri: Uri,
    context: Context
): List<Contact> = withContext(Dispatchers.IO) {
    // Define the columns we want to retrieve from the ContactPicker ContentProvider
    val projection = arrayOf(
        ContactsContract.Contacts._ID,
        ContactsContract.Contacts.DISPLAY_NAME_PRIMARY,
        ContactsContract.Data.MIMETYPE, // Type of data (e.g., email or phone)
        ContactsContract.Data.DATA1, // The actual data (Phone number / Email string)
    )

    val results = mutableListOf<Contact>()

    // Note: The Contact Picker Session Uri doesn't support custom selection & selectionArgs.
    context.contentResolver.query(sessionUri, projection, null, null, null)?.use { cursor ->
        // Get the column indices for our requested projection
        val contactIdIdx = cursor.getColumnIndex(ContactsContract.Contacts._ID)
        val mimeTypeIdx = cursor.getColumnIndex(ContactsContract.Data.MIMETYPE)
        val nameIdx = cursor.getColumnIndex(ContactsContract.Contacts.DISPLAY_NAME_PRIMARY)
        val data1Idx = cursor.getColumnIndex(ContactsContract.Data.DATA1)

        while (cursor.moveToNext()) {
            val contactId = cursor.getString(contactIdIdx)
            val mimeType = cursor.getString(mimeTypeIdx)
            val name = cursor.getString(nameIdx) ?: ""
            val data1 = cursor.getString(data1Idx) ?: ""

            // Determine if the current row represents an email or a phone number
            val email = if (mimeType == Email.CONTENT_ITEM_TYPE) data1 else null
            val phone = if (mimeType == Phone.CONTENT_ITEM_TYPE) data1 else null

            // Add the parsed contact to our results list
            results.add(Contact(contactId, name, email, phone))
        }
    }

    return@withContext results
}

Swift is designed to be the language you reach for at every layer of the software stack. Whether you’re building embedded firmware, internet-scale services, or full-featured mobile apps, Swift delivers strong safety guarantees, performance control when you need it, and expressive language features and APIs.

Swift 6.3 makes these benefits more accessible across the stack. This release expands Swift into new domains and improves developer ergonomics across the board, featuring:

• More flexible C interoperability
• Improvements to cross-platform build tooling
• Improvements for using Swift in embedded environments
• An official Swift SDK for Android

Read on for an overview of the changes and next steps to get started.

Language and Standard Library

C interoperability

Swift 6.3 introduces the @c attribute, which lets you expose Swift functions and enums to C code in your project. Annotating a function or enum with @c prompts Swift to include a corresponding declaration in the generated C header that you can include in your C/C++ files:

@c
func callFromC() { ... }

// Generated C header

void callFromC(void);

You can provide a custom name to use for the generated C declaration:

@c(MyLibrary_callFromC)
func callFromC() { ... }

// Generated C header

void MyLibrary_callFromC(void);

@c also works together with @implementation. This lets you provide a Swift implementation for a function declared in a C header:

// C header

void callFromC(void);

// Implementation written in Swift

@c @implementation
func callFromC() { ... }

When using @c together with @implementation, Swift will validate that the Swift function matches a pre-existing declaration in a C header, rather than including a C declaration in the generated header.

Module name selectors

Swift 6.3 introduces module selectors to specify which imported module Swift should look in for an API used in your code. If you import more than one module that provides API with the same name, module selectors let you disambiguate which API to use:

import ModuleA
import ModuleB

let x = ModuleA::getValue() // Call 'getValue' from ModuleA
let y = ModuleB::getValue() // Call 'getValue' from ModuleB

Swift 6.3 also enables using the Swift module name to access concurrency and String processing library APIs:

let task = Swift::Task {
  // async work
}

Performance control for library APIs

Swift 6.3 introduces new attributes that give library authors finer-grained control over compiler optimizations for clients of their APIs:

• Function specialization: Provide pre-specialized implementations of a generic API for common concrete types using @specialize.
• Inlining: Guarantee inlining — a compiler optimization that expands the body of a function at the call-site — for direct calls to a function with @inline(always). Use this attribute only when you’ve determined that the benefits of inlining outweigh any increase in code size.
• Function implementation visibility: Expose the implementation of a function in an ABI-stable library to clients with @export(implementation). This allows the function to participate in more compiler optimizations.

For a full list of language evolution proposals in Swift 6.3, see the Swift Evolution dashboard.

Package and Build Improvements

Swift 6.3 includes a preview of Swift Build integrated into Swift Package Manager. This preview brings a unified build engine across all supported platforms for a more consistent cross-platform development experience. To learn more, check out Preview the Swift Build System Integration. We encourage you to try it in your own packages and report any issues you encounter.

Swift 6.3 also brings the following Swift Package Manager improvements:

• Prebuilt Swift Syntax for shared macro libraries: Factor out shared macro implementation code into a library with support for swift-syntax prebuilt binaries in libraries that are only used by macros.
• Flexible inherited documentation: Control whether inherited documentation is included in command plugins that generate symbol graphs.
• Discoverable package traits: Discover the traits supported by a package using the new swift package show-traits command.

For more information on changes to Swift Package Manager, see the SwiftPM 6.3 Release Notes.

Core Library Updates

Swift Testing

Swift Testing has a number of improvements, including warning issues, test cancellation, and image attachments.

• Warning issues: Specify the severity of a test issue using the new severity parameter to Issue.record. You can record an issue as a warning using Issue.record("Something suspicious happened", severity: .warning). This is reflected in the test’s results, but doesn’t mark the test as a failure.
• Test cancellation: Cancel a test (and its task hierarchy) after it starts using try Test.cancel(). This is helpful for skipping individual arguments of a parameterized test, or responding to conditions during a test that indicate it shouldn’t proceed.
• Image attachments: Attach common image types during a test on Apple and Windows platforms. This is exposed via several new cross-import overlay modules with UI frameworks like UIKit.

The list of Swift Testing evolution proposals included in Swift 6.3 are ST-0012, ST-0013, ST-0014, ST-0015, ST-0016, ST-0017, and ST-0020.

DocC

Swift 6.3 adds three new experimental capabilities to DocC:

• Markdown output: Generate Markdown versions of your documentation pages alongside the standard rendered JSON covering symbols, articles, and tutorials. Try it out by passing --enable-experimental-markdown-output to docc convert.
• Per-page static HTML content: Embed a lightweight HTML summary of each page — including title, description, availability, declarations, and discussion — directly into the index.html file within a <noscript> tag. This improves discoverability by search engines and accessibility for screen readers without requiring JavaScript. Try it out by passing --transform-for-static-hosting --experimental-transform-for-static-hosting-with-content to docc convert.

• Code block annotations: Unlock new formatting annotations for code blocks, including nocopy for disabling copy-to-clipboard, highlight to highlight specific lines by number, showLineNumbers to display line numbers, and wrap to wrap long lines by column width. Specify these options in a comma-separated list after the language name on the opening fence line:

  ```swift, nocopy
  let config = loadDefaultConfig()
  ```
  

   ```swift, highlight=[1, 3]
  let name = "World"       // highlighted
  let greeting = "Hello"
  print("\(greeting), \(name)!")  // highlighted
  ```
  

  ```swift, showLineNumbers, wrap=80
  func example() { /* ... */ }
  ```
  

  DocC validates line indices and warns about unrecognized options. Try out the new code block annotations with --enable-experimental-code-block-annotations.

Platforms and Environments

Embedded Swift

Embedded Swift has a wide range of improvements in Swift 6.3, from enhanced C interoperability and better debugging support to meaningful steps toward a complete linkage model. For a detailed look at what’s new in embedded Swift, see Embedded Swift Improvements coming in Swift 6.3.

Android

Swift 6.3 includes the first official release of the Swift SDK for Android. With this SDK, you can start developing native Android programs in Swift, update your Swift packages to support building for Android, and use Swift Java and Swift Java JNI Core to integrate Swift code into existing Android applications written in Kotlin/Java. This is a significant milestone that opens new opportunities for cross-platform development in Swift.

To learn more and try out Swift for Android development in your own projects, see Getting Started with the Swift SDK for Android.

Thank You

Swift 6.3 reflects the contributions of many people across the Swift community — through code, proposals, forum discussions, and feedback from real-world experience. A special thank you to the Android Workgroup, whose months of effort — building on many years of grassroots community work — brought the Swift SDK for Android from nightly previews to an official release in Swift 6.3.

If you’d like to get involved in what comes next, the Swift Forums are a great place to start.

Next Steps

Try out Swift 6.3 today! You can find instructions for installing a Swift 6.3 toolchain on the Install Swift page.