AI-Powered Virtual Assistants

AI agents just replaced an entire SDR team and most sales leaders are still debating whether to hire more reps The math isn't even close anymore. I'm watching companies deploy AI agents that research prospects, write personalized outreach, handle initial responses, and book qualified meetings. All day. Every day. Without coffee breaks or commission negotiations. Meanwhile, sales leaders are posting job openings for 20-person SDR teams. Here's what's happening right now AI agents cost $2K monthly and generate 200+ personalized touchpoints daily Traditional SDR costs $80K+ annually and manages 15-20 conversations weekly AI agents book 40-60 qualified meetings per month Average SDR books 2-4 meetings monthly Your competition isn't hiring more SDRs. They're automating volume prospecting and repositioning their humans for high-value activities. The new SDR role looks completely different → AI strategy manager → Complex conversation specialist → Key account relationship builder → Quality control for automated outreach These aren't basic chatbots sending LinkedIn spam. AI agents understand context, industry nuances, and buyer psychology better than most junior reps. The companies waiting for perfect AI solutions will find themselves competing against teams that automated 70% of sales development six months ago. This isn't coming next year. It's happening now while you're interviewing candidates. Ready to future-proof your sales development approach? Check out The Innovative Seller for the framework on evolving your team before the market forces the change — ♻️ Repost this if you're seeing the shift Follow for more AI and sales insights

Gmail’s AI email assistant writes like a committee of lawyers designed it. Pete Koomen’s recent post Horseless Carriages explains why: developers control the AI prompts instead of users. In his post he argues that software developers should expose the prompts and the user should be able to control it. He inspired me to build my own. I want a system that’s fast, accounts for historical context, & runs locally (because I don’t want my emails to be sent to other servers), & accepts guidance from a locally running voice model. Here’s how it works: 1. I press the keyboard shortcut, F2. 2. I dictate key points of the email. 3. The program finds relevant emails to/from the person I’m writing. 4. The AI generates an email text using my tone, checks the grammar, ensures that proper spacing & paragraphs exist, & formats lists for readability. 5. It pastes the result back. Here are two examples : emailing a colleague, Andy (https://lnkd.in/gtjt3BPp), & a hypothetical founder (https://lnkd.in/gDwM4f22). Instead of generics, the system learns from my actual email history. It knows how I write to investors vs colleagues vs founders because it’s seen thousands of examples. The point isn’t that everyone will build their own email system. It’s that these principles will reshape software design. - Voice dictation feels like briefing an assistant, not programming a machine. - The context layer - that database of previous emails - becomes the most valuable component because it enables true personalization. - Local processing, voice control, & personalized training data could transform any application, not just email, because the software learns from my past uses We’re still in the horseless carriage era of AI applications. The breakthrough will come when software adapts to us instead of forcing us to adapt to it. Centered around a command line email client called Neomutt (https://neomutt.org/). The software hits LanceDB, a vector database with embedded emails & finds the ones that are the most relevant from the sender to match the tone. The code is here (https://lnkd.in/gZ-AaAWa).

Anthropic just shipped Skills, Microsoft 365 integration, and enterprise search for Claude. After talking to dozens of enterprise companies this year, I think they're solving the right problems. 💰Context tax is killing enterprise AI adoption. Most AI tools require you to manually gather information before asking useful questions. You're copying emails, uploading documents, explaining organizational context. The AI might be smart, but you're doing all the integration work. Claude's Microsoft 365 connector changes this. Direct access to SharePoint, Outlook, Teams, and OneDrive means the AI already knows what your organization knows. Ask about Q3 strategy, and it pulls from the actual discussions, documents, and decisions. They also launched Skills — reusable instruction bundles that work across Claude's web app, API, and command-line tool. Think of these as expertise packages—instructions, scripts, and resources Claude loads on-demand. And lastly, the new Enterprise search is a shared project that searches multiple connected tools simultaneously. One query pulls information from HR docs in SharePoint, email discussions in Outlook, and team guidelines from various sources—then synthesizes it into a single answer. Model providers like Anthropic and OpenAI are realizing that enterprise AI needs to be operational, not just conversational. Less chatbot, more sidekick that accesses your actual systems and takes action.

Can You Tell the Difference Between AI and Human-Written Emails? I had the opportunity to serve as an expert judge for a Washington Post feature by Geoffrey Fowler, evaluating five emails written by different AI tools—Claude, DeepSeek, ChatGPT, Copilot, and Gemini—plus one from a human writer. The challenge? Figure out which email was the best. The results? Claude won - (it even beat the human writer!) Here’s why—and what it tells us about the future of AI-powered communication: -AI Emails Are Improving—But Not All Are Created Equal Each AI-generated email had a distinct style. Some were overly formal, others were robotic, and a few completely missed the mark. Claude’s writing stood out for being the most natural, structured, and persuasive. But here’s the catch: AI-generated emails don’t all sound the same. Each tool has its own strengths and weaknesses:   •  Claude excelled at clarity, nuance, and sounding the most human.   •  ChatGPT was engaging but sometimes too wordy.   •  Copilot was direct but lacked warmth.   •  Gemini and DeepSeek struggled with context and precision. The takeaway? Choosing the right AI assistant matters. - The Best AI Emails Feel Personal—Without Overdoing It One of the biggest issues with AI-generated emails is tone mismatch. Many sound overly polished—so much so that they feel fake or even manipulative. For professionals using AI to write emails, this is key: Your AI assistant should enhance your tone, not erase it. - AI Is Great at Writing—But Terrible at Judgment One thing was clear from judging these emails: AI is brilliant at sentence structure but still struggles with discernment. Some AI tools misread the email’s context, making unnecessary recommendations. None of the AI-generated emails fully adapted to the emotional nuances of the request. This reinforces what I always tell executives: AI can help draft, but humans must decide. Before sending an AI-assisted email, ask yourself: -Does this actually answer the question? -Does it sound like me? -Would I feel good receiving this? The best professionals don’t just use AI—they train it to reflect their judgment, communication style, and emotional intelligence. Can you detect the difference between AI-written and human emails? Share your thoughts in the comments! (and full article in comments!)

Last week, the AI safety world caught its breath as security researchers demonstrated how an LLM could be tricked by a fake email into opening malicious links, all without human intervention. That’s right: your AI assistant, heralded as your productivity savior, might just become your most polite cybercriminal. While we’ve long accepted that humans are the weakest link in cybersecurity, it seems we’ve successfully outsourced that weakness to our machines. Progress? AI systems are now parsing emails, browsing the web, summarizing documents—and doing it all with the trust level of a golden retriever and the attention span of a squirrel. Combine that with prompt injection attacks that still fly under the radar, and we’re not far from an LLM confidently summarizing “Prince of Nigeria” emails as legitimate business opportunities. The takeaway? Governance is no longer a checkbox—it’s an arms race. If your AI policy consists of “don’t be evil,” you might want to update it to “don’t be gullible.” In the meantime, CISOs should start asking themselves: • Who’s vetting what our AI tools can access? • What’s our policy for AI-generated actions? • And most importantly: Can our AI tell the difference between a phishing attempt and a meeting request from Steve in accounting?

AI Agents rewrote how we win clients in 2025. After 250+ hours of research, here's what I found: They took over many GTM use-cases 👇 - Segmenting/qualifying leads on autopilot. - Conduct deep prospect research. - Personalize outreach at scale. - Build prospecting workflows. ... and more. Agents typically follow this logic: A. Figure out actions to be taken. B. Perform set of actions automously. C. Generate outputs + describe what they did. The best ones focus on verticalized use-cases: 1. Agent builders ↳ They help you craft custom agentic workflows. Examples: - Relevance AI - Relay.app - n8n - Dify - Taskade - wordware (YC S24) The best way I found to utilise these is to review manual, time-consuming, processes and automate them through the above platforms. 2. Scraping Agents ↳ Help you extract online data at scale. - Clay's AI Agent - Browserbase - Gumloop - ZenRows - Firecrawl - Apify The best way I found to use them is to think: "what's a data point that, if was true about my target, would mean they're a perfect fit for my offering?". Get creative with it... these agents help you get hard-to-find data. 3. MCP Servers ↳ Enable AI Agents to take real-world actions by connecting to your tool stack. E.g: You connect your LLM (e.g: Claude) to your Slack, Notion, Calendar, your Gmail, Stripe & even your Sales Engagement tools. You can now perform tasks like: - Rescheduling and cancelling meetings by prompting Claude. - Monitor your mailbox by asking Claude if you missed any emails. But also... Build entire prospecting campaigns by connecting your LLM to Clay, Instantly & other tools... without having to be in the interfaces of the tools. This means controlling all your work through one chat interface. MCPs are the glue between your LLM & your tech stack: Examples include: - Docker, Inc - Pipedream - Composio - Zapier 4. AI SDRs ↳ They automate cold email & LinkedIn prospecting. - Artisan - Valley - 11x - AiSDR - Topo (YC W24) - Jason AI by Reply Everyone criticises them, but the ones I've seen at play (Artisan, Valley) are generating very decent outputs. Still, make sure to review these manually. 5. GTM Co-Pilots ↳ Help you run human-crafted, complex GTM workflows. - Instantly.ai - Unify - Copy.ai - Bardeen - Common Room This category is broad, as all have a different approach. But, as an example, Instantly.ai's new copilot allows you to build entire prospecting campaigns (including building lists, writing sequences, and automating sending) within their platform by writing a few prompts. 6. Sales Agents ↳ Gather insights from your CRM / Call transcripts to help you close deals. Examples include: - Attention - Attio - Gong - Momentum.io 7. Research Agents ↳ They help you perform deep account research at scale. - Claygent - Linkup - Airtop - Exa - Tavily - Perplexity 8. List building Agents ↳ Help you build prospect lists. - Instantly.ai - Openmart - Exa P.S: Which platform did I miss? 👇

The traditional SDR role is dead. Not in 10 years. But today. Here's why:   Most teams are using the old model.   Hire junior reps to blast cold emails. Book demos. Burn out in 12 months.   It’s slow. Expensive. And it doesn’t scale.   But a shift is happening:   AI isn’t just making SDRs faster.   It’s replacing them.   The best GTM teams don’t hire more reps. They build systems that do what reps can’t:   - Spot high-intent leads on LinkedIn - Enrich contact data in seconds - Personalize every message - Reach out minutes after the signal - Nurture across LinkedIn, email, and SMS   We’ve built that exact system at Atticus. Here’s what it looks like:   1️⃣ Signals + Lead Scoring   → Tools: Valley, Clay, RB2B → Auto-detect profile views, site visits, competitor follows → Score leads against our ICP in real time   This gives us a live queue of warm leads, before most teams even know who to contact.   2️⃣ Enrichment + Research   → Tools: Clay, Prospio → Pull job titles, emails, hiring activity, and funding rounds → Auto-generate talking points and likely pain points   No guessing. Just hyper-relevant context, pulled and packaged instantly.   3️⃣ Outreach + Routing   → Tools: Valley, Smartlead, HeyReach → Send personalized DMs, cold emails, and social touches → Route hot leads to the AE or founder with zero friction   The system does the work. Our team only steps in once a conversation’s started.   4️⃣ Human-Like Follow-Up   → Tools: Custom AI agents trained on past conversations → Handle replies, ask qualifying questions, and direct to demos → Handover to AE via DM, email, or SMS when ready   Every reply gets a response. Every signal gets actioned.   No leads slip through the cracks.   The results?   → Replaced 2 full-time SDRs → Booked 30–50 SQLs/month → £0 in ad spend → 3–5x reply rates compared to manual outreach   The future of outbound isn’t headcount.   It’s systems. P.S Peep the screenshot below on a recent outbound campaign - 33% response rate - 688 messages, 54 leads and 9 meetings booked. I believe the acceptance rate was lower because a value prop was given in the first message. We currently aren't able to empty connect on software.

Everyone talks about using AI for writing. I use Claude to run my day. It’s not a tool. It’s an operations partner—if you give it the right prompts. Here’s exactly how I use Claude as my assistant (connected to Gmail, Drive, and Calendar): 1. Morning Briefing Prompt Start the day with clarity. “Check my calendar, unread emails, and recent docs. Summarize today’s meetings with prep notes. Pull any open loops or tasks from emails. Suggest a time-blocked plan for deep work + admin. Flag anything urgent or out of alignment.” I open Claude before I open my email. 2. Pre-Meeting Prep Prompt No more last-minute scrambling. “I have a meeting with [Name] about [Topic]. Pull key context from emails, docs, and last calendar invite. Extract action items from last call. Draft talking points and 3 smart questions to ask.” Perfect for client calls or collabs. 3. Research & Synthesis Prompt Working on a project? Claude becomes your researcher. “I’m working on [project]. Pull relevant threads from Gmail. Scan docs with [keyword] and summarize insights. Build a timeline of progress + open items. Draft a quick project update I can send or post.” This alone has saves me 3 hours a week. 4. Workspace Organization Prompt Your brain, but with folders. “Find all docs related to [project]. Suggest categories or themes. Create a folder/tag structure that makes sense. Highlight outdated files or duplicated info. Build a cheat sheet with links + purposes.” Perfect if your Google Drive looks like a tornado. 5. Smart Inbox Prompt Catch up without the chaos. “Find unread emails from VIP contacts. Summarize key threads and flag what’s urgent. Draft quick replies where possible. Link any emails to related docs or calendar events. Build a follow-up plan so nothing slips.” It’s triage for your inbox—with logic. Claude isn’t just for content. It’s for operations, decisions, and daily momentum. Want more tips like this? Join 3,400+ readers of 9-To-Thrive → https://lnkd.in/gXMzXweK

Last week Wesley tried to renew his car insurance. He thought it’d take 10 minutes. It took 2 days, 3 back-and-forth calls, and an email thread with someone who had to “check with another team.” All they needed was his policy details to send a quote, and a payment link. The rep was polite, but the process was broken. Not the tech itself, but the glue work between systems, approvals, and processes. The hidden cost in insurance isn’t just fraud, it’s friction. That’s where CozmoX AI (YC W22) AI comes in. We’ve built Voice AI Employees that own it end to end for insurance companies. Here’s how it plays out technically for top insurers: 1. Caller Interaction (Telephony Layer) We integrate natively with SIP or cloud telephony (Twilio, Genesys, Avaya, etc.) to receive inbound or make outbound calls. Our AI Employee answers instantly with natural-sounding speech, no IVR menus. 2. Intent Recognition (NLP Layer) Using deep context windowing (vLLM + custom NLU), we detect whether the caller wants to file a claim, ask about a policy, or renew—no rigid keyword matching needed. 3. Contextual Memory (Session + External Memory Layer) Our AI Employee remembers. It pulls customer info live from CRMs (Salesforce, etc.), policy systems, or claims platforms (Guidewire, Duck Creek) via secure APIs. 4. Action Execution (RPA/API Layer) Once intent is confirmed, the AI triggers backend actions: Files FNOL Fetches policy docs Updates payment status Starts renewal workflows All done via REST/SOAP APIs or RPA if systems are legacy. 5. Real-Time CRM Sync (Logging Layer) Everything is logged: transcript, summary, outcome, next steps compliance ready and analytics-friendly. This isn’t a chatbot with a voice. It’s a full-stack operational AI built for regulated, high-stakes, high-volume industries like insurance. And the impact with an insurance aggregator we are working with - 80–90% automation of inbound/outbound calls - 50% drop in average handling time - 2x boost in customer satisfaction - Full traceability with structured logs + consent capture We’re not replacing people. We’re removing the repetitive glue work that stops them from working at the top of their license. If your team is still stitching together CRMs, call scripts, and manual workflows - we should definitely talk.

Excited to share the successful launch of a GenAI project we began in late 2023, during the early days of enterprise generative AI adoption. Despite numerous challenges along the way, our persistence and collaborative effort led to an innovative solution. Working alongside the exceptional Verisk team (Sundeep Sardana, Malolan Raman, Maitri Shah) and my Amazon Web Services (AWS) colleagues Alex Oppenheim ☁️ and Tarik Makota we were able to bring this vision to life. We built Verisk's new generative AI-powered companion for their Mozart insurance policy management platform. This solution transforms the policy review process, reducing what traditionally took days or weeks into just minutes by intelligently comparing and summarizing policy document changes. Want to dive deeper into our journey? Check out our detailed blog post covering the architecture, challenges, and key learnings: https://lnkd.in/dnTrnGpz #GenerativeAI #AWS #Innovation #InsurTech #CloudComputing