Cold outreach at volume is a numbers game, and volume needs people. Sherry needed close to seven figures in sponsorships for her AI conference, with no sales team.So she automated it. Manus built a list of 500 companies that had already sponsored comparable AI conferences. She filtered for the roles that field sponsorship inbound, then pointed a browser agent at her LinkedIn to send requests one by one, mimicking a human instead of hitting an API that would get her blocked. It ran overnight. Replies came around 10 to 15 percent, with no personalisation at all.

In this episode:

* Why she raised seven figures in sponsorships without a single salesperson

* How a research agent produced a 500 company sponsor list from past conference exhibitors

* Why she removed personalisation from her messages, including the recipient’s name

* How browser automation gets around LinkedIn’s limits when APIs get blocked

Why mutual connections beat personalised copy on reply rate

* How she filtered for growth, business development, developer relations, and founder titles

* Why the agent has to be painfully slow, and why that means running it overnight

* What the whole thing actually cost compared to a virtual assistant at 7 to 10 dollars an hour

* Why AI outreach is only as good as how clearly you’ve defined who you wantI

Timestamps:

(00:00) Introduction

(00:49) Seven figures in sponsorships, no sales team

(02:02) How she did it before automating

(03:19) Why she cut the personalisation

(04:41) Building a 500 company list with Manus

(06:05) Demo: researching sponsors live

(09:59) Why the agent needs its own computer

(11:10) Getting around LinkedIn’s automation limits

(13:22) Real messages, sent by AI

(14:36) What the credits cost

(16:03) The 15% reply rate

(17:15) The wrong name mistake

(18:40) Get clear before you automate

(20:08) Agency and a half million dollar deposit

Resources & Links:

* Manus

* Codex

* AI Engineer Singapore

Connect with Sherry Jiang:

* LinkedIn

* Peek

* X

Most founders know they should post more short-form content. The hard part is finding what works and staying consistent without a full UGC team.

Ayush had an extra constraint: his customers are in the US, but TikTok is banned in India. So he built a workaround.

In this episode of Shipping with AI, he walks through the AI UGC workflow he uses to test TikTok content from India. He created Kate, an AI influencer for his dictation app, then used Fastlane.ai and Claude to generate content, find proven formats, and post through warmed up US accounts.

In this episode:

* Why TikTok being banned in India does not stop Ayush from testing TikTok content

* How an AI influencer generated 25,000 organic views in one monthThe four narrow sub-agents behind the analyst, and why narrow scope beats one agent doing everything

* How Fastlane.ai uses warmed up TikTok accounts to publish into the US market

* Why people accounts can work better than brand accounts for UGC-style content

* Why Ayush treats the account as a testing ground before spending on paid ads

* How the workflow replaces a creator, editor, researcher, and social media operator

* Why AI content quality matters more than whether the audience knows it is AI

Timestamps:

(00:00) Introduction

(00:43) Why TikTok is hard to access from India

(01:53) Running TikTok without ever using TikTok

(03:33) Meet Kate, the AI UGC influencer

(05:25) 25,000 organic views from 54 posts

(06:31) How Fastlane.ai finds proven TikTok formats

(08:03) Using warmed up TikTok accounts in the US

(10:01) Creating the AI influencer with Claude

(13:06) Why use a person account instead of a brand account

(14:31) Testing hooks before scaling content

(15:38) Creating AI images and videos for Kate

(19:36) Using Blitz to pick viral content formats

(20:19) Turning a proven format into a VoIP promo

(22:13) What content formats worked best

(23:41) Scheduling 8 to 10 days of content in 2 hours

(24:38) Tracking branded searches and conversions

(26:27) Cost breakdown of the AI TikTok workflow

(27:50) Breaking even with one or two customers

(28:56) Does AI content cap out on TikTok

(30:06) Why this replaces a full UGC production team

(31:54) The future of AI-run marketing workflows

(34:40) Final thoughts and wrap-up

Resources & Links:

* Claude

* Fastlane

Connect with Ayush Chaturvedi:

* LinkedIn

* Voibe

Discover how a seed-stage investor turned a slow, manual research process into an AI analyst he runs from a single Slack message. This deep dive walks through why most funds won’t let AI near their confidential deal documents and why he did it anyway, the exact architecture that stops the agent from inventing numbers on financial data, and a live demo of the whole thing pulling a full company brief from four connected systems in seconds.

In this episode:

* Why trusting AI with confidential data is really a question about the vendor, not the technology

* How he built the first working version over a weekend, and why it is never actually finished

* The four narrow sub-agents behind the analyst, and why narrow scope beats one agent doing everything

* The rule that keeps it honest: clickable sources, a human review step, and treating the agent like an intern

* When to call a plain function instead of letting the model use judgment, so it stops making things up

* Why the near future is agent to agent, not human to human, and what that changes for founders

* How to build your own, whether you want a frontier model’s harness or full open-source control

Timestamps:

(00:00) Introduction

(00:40) The problem: Portfolio answers scattered across systems

(02:49) Why let AI near confidential deal documents at all

(04:11) Do you actually trust OpenAI and Anthropic with your data?

(04:26) Choosing a harness and locking it down with Tailscale

(07:36) Building the first version over a weekend

(08:47) Live demo: a full company brief from one Slack DM

(10:48) Why the near future is agent to agent

(12:30) Verification: clickable sources and the intern rule

(15:14) The architecture: Claude Code and the Agent SDK

(18:24) Compounding org knowledge with ByteRover

(21:18) The four sub agents and the data cleanup grind

(24:43) When to call a function instead of trusting LLM judgment

(28:44) Loops, goals, and orchestrator burnout

(29:47) Advice for building your own AI analyst

(32:51) Turn paranoia into verification, then just build

(34:06) Wrap up

Resources & Links:

* Claude

* Attio’s docs built for AI agents: https://attio.com/llms.txt

Connect with Binh Tran:

* LinkedIn

* avv.co

Discover how a CTO turned the messy problem of vibe coded apps into a single Claude skill that tells you whether your app is actually safe to ship. This deep dive walks through why asking AI “is this production ready” burns through hundreds of thousands of tokens and still gives you the wrong answer, the seven to ten things a CTO actually checks before an app goes live, and two real internal tools audited on screen, including the moment the skill catches a leaked secret key and app logic sitting wide open on the client side.

In this episode:

* Why maintainability matters as much as security the moment real people start using what you built

* What a non-engineer simply cannot see when they look at a working app, and where it breaks

* Why just asking AI if something is production ready burns 200 to 250K tokens and changes its answer the next day

* How an app ends up split across Vercel, Railway, and AWS when the AI keeps improvising fixes

* The CTO’s checklist: tech stacks, backing resources, code repos, secret scanning, libraries, and schema

* A single-prompt Claude skill that ranks every issue by how soon it will bite you, then hands back a fix list

* Where the skill still traps you in a fix-then-break loop, and the exact moment to bring in an engineer

Timestamps:

(00:00) Introduction

(01:07) The problem: your team can build apps but can’t tell what’s safe to ship

(02:26) Why maintainability matters, not just security and safety

(03:35) What a non-engineer can’t see when they look at working code

(05:27) Why just asking AI “is it production ready” fails

(05:44) The token burn: 200 to 250K per run, and answers that change daily

(06:37) How an app ends up split across Vercel, Railway, and AWS

(07:16) The CTO’s checklist for what makes an app production ready

(08:09) Tech stacks, backing resources, code repos, and secret scanning

(10:19) Live demo: auditing a simple work redesign app

(11:47) The prototype a non-coder built in minutes

(12:43) How the skill runs the audit on the app

(13:14) The biggest risk it caught: all logic on the client side

(14:25) Second example: the asset tool that replaced a SaaS

(17:08) Walking through the audit results and the fixes

(17:51) One prompt to fix a leaked secret key and more

(19:31) Running the skill with Claude in thinking mode

(20:02) Which model and thinking mode to use for fixes

(20:40) How to tell if you’ve overbuilt or burned too many tokens

(22:13) Inside the prompt: sanity checks and severity ranking

(23:14) How non-engineer proof the skill really is

(23:57) Does it explain the risks in plain language?

(24:53) Has the skill ever missed anything significant?

(25:30) The fix-then-break loop and how to get out of it

(27:56) What maintainability actually means

(29:25) Does it work with only the prompt history?

Resources & Links:

* Claude

* https://github.com/impressai/vibe-check-skill

* https://www.12factor.net/

Connect with Dr. Vaisagh VT:

* LinkedIn

* Impress.ai

Arjun Paul runs Zoko, helping e-commerce brands handle WhatsApp at scale. One of his customers gets 82,000 messages a month — three specialized bots (one for order tracking, one for FAQs, one for sales) now handle 75% of those chats before a human ever sees them. The team finally has time to work on the chats that actually make money.

This episode walks through the three-bot architecture, the pricing logic that justifies the cost, and why none of this worked six months ago.

In this episode:

* Why WhatsApp converts at ~30% vs ~3% on a website — and why that creates a support flood worth solving

* The breakdown of 82,000 monthly messages: 40–60% asking “where is my order?”, another 40% repeated FAQs, 20% sales

* Three specialised bots — Wismo (order tracking), Guru (FAQ), Sello (sales) — and why one-task agents beat general ones

* What actually changed six months ago: the base models got good enough, not their fine-tuning

* The headline numbers: Wismo deflecting 52% of all chats with 93% resolution

* How Zoko prices the bots — hours saved × local labor rate × typing speed, not tokens

* The free-trial-then-bill playbook: run it free for three months, customers can’t go back to humans after that

* Guardrails: the handover protocol, refusing off-topic queries, the private-note transparency loop

* What good looks like for FAQ bots: business owners seed ~500 questions, resolution climbs from 70% toward 90%

* The next frontier: bots that actually sell — taking orders in Urdu, sending checkout links, doing outbound

* Setup for Shopify merchants: one click to enable Wismo

Timestamps:

* (00:00) — Cold open: 82,000 messages, three bots, 75% handled without a human

* (00:39) — Paint the old world: what was broken on WhatsApp support

* (00:54) — Why support floods WhatsApp: 30% conversion vs 3% on websites

* (02:30) — Bucketing incoming chats with an LLM; “where is my order?” is 40–60% of messages

* (03:30) — Customers check order status ~4 times before delivery, across every market

* (04:10) — The three-bot architecture: one agent, one task, one outcome

* (05:18) — The 82,000-messages-per-month customer

* (06:49) — Why specialized bots win: 40% Wismo, 40% Guru, 20% sales

* (07:27) — What actually changed in six months: base models got better, not the fine-tuning

* (08:37) — The headline number: Wismo handled 52% of all chats, resolved 93% of them

* (10:14) — Pricing logic: hours saved × hourly rate × 45 seconds per message

* (11:50) — The free-trial close: run it on for three months, then have the pricing conversation

* (13:19) — Measuring resolution: the 80% threshold that made it sellable, two years to get there

* (16:08) — Handover protocols and refusing to answer off-topic questions

* (17:00) — Meet Guru: the FAQ bot at 70% resolution and how to push it toward 90%

* (19:33) — Measuring quality with private notes: how Zoko earns customer trust on resolutions

* (20:45) — The work the business owner has to put in: ~500 seed questions for the knowledge base

* (23:08) — What’s next: bots that sell, take orders in Urdu, do outbound marketing

* (25:53) — Setup: one click to enable Wismo and Guru inside Zoko

* (26:46) — Why Shopify: connecting orders and catalog into the WhatsApp screen

* (29:56) — The benchmark customer: 45% Wismo + 30% Guru = 75% of chats handled by AI

Resources & Links:

* Zoko

* Shopify

Connect with Arjun Paul

* LinkedIn

Discover how an enterprise sales founder uses Claude to turn raw call recordings into tailored, country-specific proposals in a fraction of the usual time. This deep dive walks through the connectors he set up, the prompt he never actually wrote by hand, the human review gates that keep client data safe, and the lessons from iterating the same workflow more than ten times.

In this episode:

* Why generic templates fail for enterprise proposals, and where customization actually moves the deal (currency, country-specific regulations)

* Building a Claude automation that turns a sales call into a draft proposal

* How the setup connects Fathom recordings, the Attio CRM, and his existing proposals

* The role of outline approval, draft-only outputs, and a human review checklist

* Getting client consent before recording, and protecting sensitive information

* How iteration took proposal time from over two hours down to roughly thirty minutes

* Why the bar for enterprise sales keeps rising, and where human relationships still decide deals

Timestamps:

* (00:00) Intro: Vinayak’s sales call volume and proposal workload

* (01:01) Cutting proposal time from 2+ hours to 30 minutes with Claude

* (01:46) Why Claude became the go-to AI tool

* (02:15) Why templates don’t work: the case for customization

* (03:08) A real-world example: currency and regulatory differences by country

* (04:48) Using AI to create meaningfully different customer experiences

* (05:32) Who this workflow is for: the ideal personas

* (07:08) Live demo of the automation setup

* (08:15) The automation prompt and how it was built

* (09:57) Walkthrough of a real output: the US client deck

* (11:33) How to control what gets emphasized in proposals

* (12:14) Lessons from iterating the automation

* (13:21) The current setup, with manual review checkpoints

* (14:19) Human in the loop: keeping an approval gate

* (15:07) How much the output improved through iteration

* (16:24) What happens when recordings are incomplete

* (17:48) The review checklist, and how the time spent has dropped

* (18:15) Handling sensitive client information

* (20:35) What the next version of this looks like

* (22:14) Step by step: how to set this up yourself

* (22:47) Top tips for getting started

* (24:05) Why the bar for sales has gone up

* (24:34) The irreplaceable role of human relationships in enterprise sales

* (26:30) Competing in an AI-enabled sales landscape

* (26:56) Rahul’s story: why presentation quality wins deals

* (27:43) Wrap-up and closing remarks

* (27:48) About Superjoin

* (29:34) Sign-off

Resources & Links:

* Claude

* Attio (CRM)

* Fathom (meeting recorder)

* Notion

Connect with Vinayak Jhunjhunwala:

* LinkedIn

* X

* Superjoin

Discover how a property owner leverages AI automation to streamline booking reconciliation, enhance data accuracy, and gain peace of mind. This deep dive explores the technical workflow and benefits of building a custom AI-powered system for managing multi-channel bookings.

In this episode:

* The challenges of manual booking management across multiple channels

* Building a custom AI skill with Claude to automate data reconciliation

* How the system pulls data from Gmail, Airbnb calendar, WhatsApp, and offline bookings

* The role of conflict detection, issue flagging, and iterative refinement

* Practical insights on data validation, error handling, and decision-making

* Future possibilities: a web interface for broader accessibility

* The peace of mind and data-driven decision advantages for hosts

Timestamps:

* (00:43) How long he hosted before needing a system

* (01:02) Why daily manual logging was impossible

* (01:59) What was wrong with the property manager’s reporting: late and unverifiable

* (03:04) Why existing property apps don’t solve multi-channel, no export

* (04:56) The raw data he actually receives: WhatsApp, Gmail, iCal

* (06:30) The schematic: iCal as source of truth, then Gmail, then WhatsApp gap-fill

* (09:40) Running the demo skill and how it anonymises real data

* (11:20) Why it needs Gmail access and why scope can’t be narrowed

* (12:23) What a Claude skill is and how he built it iteratively

* (13:26) Resolving conflicts: Airbnb email beats WhatsApp as source of truth

* (13:58) Run time and why he runs it end-of-month

* (15:03) The offline-booking blind spot and how iCal catches it

* (16:15) The case for an Airbnb MCP server

* (17:13) Human in the loop, not full automation

* (17:57) How many real issues it caught, plus the tax-deduction edge case

* (19:39) The real payback: peace of mind and clean data

* (20:35) Hard ROI numbers versus other hosts’ gut feel

* (22:07) Flagged conflicts and the monthly summary

* (23:32) Excel and dashboard baked into the skill

* (25:12) Claude Code version versus a hypothetical GUI

* (26:01) The binary insight: you do it or you don’t

* (26:38) Building trust with the manager and the idea of layering in neighbourhood rates

* (27:42) Wrap-up

Resources & Links:

* Claude Code

Connect with Sharath C George:

* LinkedIn

* Twitter

* Website

In this episode, Sushant shares a detailed walkthrough of how they’ve built and optimized Hermes AI agent like called Yumiko to automate and improve lead generation and business workflows. Learn how cutting-edge tools are making it easier for startups to scale prospecting and outreach in minutes, not hours.

Main topics:

* Manual vs. automated lead sourcing: the evolution

* Setting up Hermes AI agent: step-by-step process

* Integrating tools like Google Cloud, Composio, Apify, Crust Data, and more

* How Hermes learns and improves over time through self-updating skills

* Cost considerations and infrastructure setup for startups

* Ensuring lead quality and avoiding hallucinations

* Practical steps for deploying a lead-generation AI in 20 minutes

Timestamps:

00:00 - How Sushant automated his lead sourcing process using Hermes02:08 - The pain points with manual LinkedIn prospecting and outreach03:04 - How Hermes searches, gathers, and organizes leads automatically04:11 - The importance of giving context to AI agents for precision05:15 - Limitations of tools like Apollo and manual lead vetting06:38 - Setting up virtual machines on Google Cloud easily08:15 - How Hermes learns and refines its lead-finding skill over time09:01 - Adding proactive lead alerts and social post monitoring10:21 - The orchestration engine powering multiple AI agents working together11:44 - Demonstration of Yumiko, the AI executive assistant, in action12:48 - Managing meetings, contacts, and data without manual input14:10 - Tools like Composio, Crustdata, and EXA for enriched lead data15:58 - Self-updating skills make Hermes more accurate over time16:36 - Common mistakes and safety guardrails during automation18:22 - How AI agents can send emails, but with safeguards20:27 - Infrastructure setup: virtual machine configurations and APIs23:00 - Connecting Hermes to messaging gateways like Telegram and Slack24:34 - Cost analysis: infrastructure + API credits, budget-friendly options34:44 - Ensuring lead authenticity and avoiding hallucinations36:24 - How simple and accessible deploying AI-based lead gen has become38:13 - Final tips for non-technical founders to set up in 20 minutes39:02 - The future of AI agents acting as virtual team members

Resources & Links:

* Hermes

* Composio

* Crust Data

* Exa Search API

* Apollo

* SimpleClaw (easy setup virtual machine service)

Connect with Sushant:

* LinkedIn

* Twitter

Note:

This methodology demonstrates that with minimal setup—roughly $50/month in cloud and API costs—any startup can automate lead sourcing within minutes, making outreach smarter and scalable. Whether you’re technical or not, the ecosystem options now support effortless deployment.

Discover how Dipesh automated his email inbox with AI agents, reducing distractions and increasing efficiency. Learn the step-by-step setup, costs involved, and potential security risks of integrating AI with your Gmail and calendar system.

Main Topics Covered:

* How Dipesh’s AI agent reads and briefs his emails three times daily

* Full live walkthrough of setting up Claus with Gmail and Google Calendar

* Prompt engineering to prevent AI from sending emails without approval

* The time investment to automate email management (~15-20 minutes setup)

* Cost considerations for using AI agents ($19–$39/month)

* Security risks including data hallucination and unauthorised access

* Strategies to safeguard sensitive information (e.g., using separate email addresses)

In this episode:

* Dipesh shares his initial email checking habits (15-18 times daily) and how automation replaced this

* Step-by-step setup process using CLAWS.app and OpenClaw framework

* How the AI filters important emails, highlights urgent actions, and syncs with Google Calendar

* Custom training of the AI using emails downloaded via Google Takeout

* The importance of setting clear instructions, e.g., “Draft only, do not send”

* How to detect and handle setup hiccups, and the tools involved

* Cost justification based on hours saved and focus gained

* Security measures: connection restrictions, minimising hallucinations, and auto-forward rules

* Practical advice for replicating the setup in under an hour

Timestamps:

00:00 — Introduction: how Dipesh stopped checking email entirely using an AI agent

00:34 — The problem: checking Gmail 15–18 times a day as muscle memory

01:50 — What he tried first: filters, rules, and unsubscribing (and why they failed)

03:27 — Live demo: the agent setup and what the daily briefings look like

07:08 — Setup time and the service he used

08:10 — Connecting Gmail and Google Calendar

08:40 — How the agent decides what’s important (without explicit instructions)

09:49 — Training the agent with historical email data via Google Takeout

10:53 — Replicating the setup live from scratch

13:47 — Scheduling briefings with plain-language instructions (9 AM, noon, 4 PM)

16:43 — The real risks of giving AI access to your Gmail

18:21 — How to prevent the agent from sending email without your approval

19:25 — Additional safety setups (separate forwarding address, local AI filtering)

21:24 — Using “skills” to lock in standing instructions

23:49 — Efficiency gains: three weeks without opening Gmail

24:19 — The cost ($19–$39/month) and whether the ROI makes sense

25:50 — How to get started in under an hour

Resources & Links:

* Klaws.app — Easiest AI automation for Gmail and Google Calendar

* OpenClaw — Framework for managing AI agents

* Google Takeout — Download all emails for training your AI

* Composio — Service used for linking Gmail with AI

Connect with Dipesh:

* LinkedIn

* Twitter

Note:

Security practices and prompt engineering are crucial to prevent unintended AI actions, hallucinations, or data breaches. Always assess the risks before giving AI access to sensitive data.

Discover how Librarian, an AI-powered executive assistant, is revolutionising productivity across various industries, from real estate to insurance. This episode explores practical use cases, integration strategies, and how to start leveraging AI tools for efficiency and peace of mind.

Main Topics:

* What is Librarian and how does it function as an AI executive assistant?

* The evolution of AI from static software to dynamic, voice-controlled workflows

* Use cases in real estate: lead management, property searches, and task automation

* How Librarian integrates with popular platforms like WhatsApp, Gmail, Google Drive, and more

* Building vertical-specific solutions: focusing on real estate and insurance

* Challenges in setup, integration, and maintaining data privacy and security

* Practical steps to get started with Librarian and maximising its features

* The broader applicability of AI assistants for lawyers, coaches, and sales teams

* Impact on peace of mind and productivity through reminders, briefs, and automation

Timestamps:

00:00 - Introduction to Librarian, an AI executive assistant for task automation00:44 - How Librarian works and interfaces with users via WhatsApp01:20 - The leap from traditional software to AI-enabled voice interactions02:13 - The origin story: from blockchain disappointment to AI revolution03:05 - The search for pain points: inefficiencies in real estate workflows03:55 - Demonstrating Librarian’s ability to search and organize information across platforms04:53 - The focus shift to verticals, especially real estate, for deeper impact05:46 - Illustrating real estate use case: driving, lead capturing, and task automation08:03 - Practical demo: managing a property lead call via WhatsApp09:29 - Invoking Librarian to share property options and reminders for clients11:10 - Capturing lead data, tasks, and scheduling follow-ups automatically12:13 - Crafting messages and updating tasks hands-free while on the move14:17 - Beyond property searches: calendar management and email automation16:42 - Workflow automation: from lead intake to automation on Slack and tickets18:29 - The value of integrations like GitHub for troubleshooting and task management20:29 - Addressing setup complexities and the value Librarian provides in stitching workflows23:19 - Ease of setup: from messaging setup to full automation with minimal effort25:22 - How Librarian enhances peace of mind and daily planning with reminders and briefs27:53 - Expanding use cases: lawyers, consultants, and sales professionals29:27 - Practical applications for insurance agents and client management30:30 - Security and privacy measures: encryption, SOC 2 certification, and data handling32:25 - How to get started: creating a free account, top features to explore33:22 - Final thoughts and call to action for productivity enthusiasts

Resources & Links:

* Librarian AI Executive Assistant

Connect with Tiago:

* LinkedIn

* Twitter

Experience the future of work now — start your free 14-day trial with Librarian and transform your productivity!