# HADEF — Full Site Content (llms-full.txt) > Concatenated full-text version of HADEF's primary public pages, intended for AI assistants that fetch long-form context. Canonical site: https://hadef.io. Short index: https://hadef.io/llms.txt. --- ## Source: Homepage (https://hadef.io/) # Landing high-value podcast guests is not a numbers game. Until now. A context engine that knows your show, your shortlist, and the angle nobody else has pressed on. It builds the pitch with you, hands you the finished package, and you land the guest looking like you did the homework. ## It's not you, it's your pitch. The truth is, it's easier than ever to write a pitch, yet harder than ever to land one. The problem is every other host's inbox is getting flooded with pitches that don't move them. We built our system to fundamentally solve this problem. By deeply understanding your show, content, and plans. Then doing the same with your guest. We find the overlap no one else pitched. What you send is specific enough to be un-ignorable. ### Transcripts from audio no one else has We extract full transcripts from podcast episodes that were never publicly transcribed. Your pitch opens with a line the guest said on an episode with 200 listeners. That's homework no LLM can replicate from training data. ### A system that remembers your show Show Brain stores your voice, your audience, your rules, your wins and losses. Every pitch is shaped by everything that came before it. ChatGPT starts from zero every single time. The gap compounds with every pitch you send. ### A pipeline, not a chat window Who you pitched, when, whether they opened it, whether they replied, when to follow up, which angles convert. That's a booking operation. ChatGPT gives you text and walks away. You're left managing the rest in your head. ## You want your work to speak for itself. Which means the pitch has to speak for the work first. You don't want to send a pitch you'd wince at a week later. So HADEF reads your show as deeply as it reads the guest. Every episode you've recorded, every recurring theme, every position you've taken on the record. Then it pulls every interview the guest has given in the last eighteen months, every article, every talk, and finds the one overlap no one else has pitched. What you send is specific enough that the reader can tell who was paying attention. ### Reads every interview before you do. Your guest's published record, podcasts, keynotes, press, becomes the brief that writes the pitch. Hey Maya , noticed hook quote in publication … ### Writes in your voice, from their quotes. A DM drafted around something they actually said, in the rhythm of how you'd write it yourself. ### You protect the people who say yes. Verified titles, current company, risk-audited questions. Guests leave willing to recommend you to the next one. ### Sends where the guest actually is. LinkedIn DM, email, X, Substack. Hadef picks the channel the guest responds to, not the one that's convenient. ## Your show's memory, captured the moment it happens. Speak it, type it, drop a link. Hadef builds your context the way you actually work, then surfaces it when you're ready to pitch the next guest. - 127 context entries - 38 guests booked - 14 topics covered - 94 Brain health - 23 Topics indexed - Ready to pitch ## Context. Pitch. Send. Three moments that map to how booking actually works. Each one earns the next. - Forbes · Mar 2026 - Bloomberg · Feb 2026 - MarketWatch · Jan 2026 > "AI agents for sales enablement were dead in the water. Until we invested in the context engineering underneath them." ### We read every interview so you don't have to. Hadef ingests your guest's public record and surfaces the one line that's worth pitching around. You don't scroll. You don't search. You arrive at the hit. > "AI agents for sales enablement were dead in the water. Until we invested in the context engineering underneath them." ### You pitch like a journalist, not a vendor. Which starts with a thread your guest has been pulling on that nobody else has picked up. You don't want to seem salesy. HADEF drafts the first sentence around a real overlap: something the guest has just said in public, a gap in your show's recent arc, a timing trigger that makes the conversation feel inevitable now. The reader opens the message, and the archive door doesn't close on you. Vendors write the other kind of email. - LinkedIn DM Picked - Gmail - +3 Hey guest name , noticed your interview in publication where you said hook quote . Would love to dig deeper… ### Hand-off to LinkedIn, in your voice. LinkedIn is where these conversations actually start. The pitch lands in the composer formatted the way you would have written it. You hit send. ## Fewer pitches. Higher hit rate. That's the point. The threshold isn't price. It's how much of the work you want to be in the room for. - 5 pitches per month - 1 show connected - Show Brain (text deposits) - 1 GB storage - QA verification on every pitch - 12 pitches per month (unused roll over, max 20) - Up to 3 shows connected - Full Brain (voice, files, auto-tags, signals) - 10 GB storage - Pitch analytics + AI situation drafting - Priority support - 25 pitches per month (unused roll over, max 40) - Unlimited shows connected - 50 GB storage - Multi-channel send (LinkedIn + email) - API access + custom integrations - Dedicated onboarding call - You tell us what a booked guest is worth - You only pay when a pitch is accepted - Hadef team runs the whole workflow - Booking targets agreed up front - Direct line to the founders All plans include the Show Brain, golden-thread pitch packages, multi-channel send, and the QA layer. Cancel any time. Annual plans save 20%. ## Why podcast hosts need help. After putting in 10,000+ hours of manually securing high-value guests for podcasts, webinars, and panel discussions, we've automated the end-to-end workflow. ### A message can fail a hundred ways. One wrong line and you've left a trail of half-read pitches in a prospect's DMs. The opportunity is done before the draft even made sense. ### Context decays faster than you can track. Keeping an AI co-pilot current on every podcast, article, LinkedIn post, and Substack your prospect ships is tedious. Finicky. Usually stale the day you use it. ### Generic LLMs run out of room. You want a pitch with nuance. LLMs without the right APIs can't reach voice-only transcripts, hit context limits, and stall on the searches that actually matter. ### Built on MCP — plug it into the agent you already use. Hadef's tools (research_prospect, get_pitch, approve_pitch) expose over MCP. Equip your agents with them, or let the app handle the whole flow. Same result. ## Questions hosts ask before they trust us with their pitches. Every pitch passes a QA layer that verifies the quote against the source URL and flags anything that doesn't check out. You see the verification before you send. If something slips through, we want to know fast. Other tools tell you they keep humans in the loop. We tell you exactly where: at send. ## Land high-value guests on your podcast today. This is what happens when the pitch is built around what your guest has actually said. ### Your next guest is one pitch away. --- ## Source: About HADEF (https://hadef.io/about-hadef) # Pitches that read like you actually did your homework. Because HADEF did. HADEF reads every public appearance your potential guest has made, finds the conversation that has not happened yet, and writes a personalized pitch in your voice. Grounded in real transcripts. Quality-scored before you ever see it. ## What HADEF actually does HADEF is a SaaS for B2B podcast hosts who interview senior leaders and currently spend hours per guest on manual research and outreach. It does one thing well: it researches a potential guest, identifies the unexplored conversation angle, and writes a personalized pitch you can send in one click via LinkedIn or email. You give HADEF a name and a LinkedIn URL. About fifteen minutes later you have a research brief built from the guest's podcast appearances, keynotes, articles, and LinkedIn posts; a proposed episode concept with five interview questions; and a pitch in your voice ready to send. Every quote referenced in the pitch has been verified against its source. The pitch comes with a 0-100 quality score so you know whether to send as-is or rework first. HADEF is not a contact database. It is not a marketplace that connects you with willing guests. It does not send messages on your behalf. It does not auto-publish anything. You stay in control of every send. The product writes; you decide. ## How the pipeline works HADEF is a sequence of five core agents, plus thirteen supporting agents that handle quality, verification, and risk. In plain language: ### 1. Research Agent Reads the public web. Pulls every podcast appearance, keynote, article, and LinkedIn post your guest has produced. Time-prioritized so the most recent material weighs most. ### 2. Transcript Agent For audio and video where no transcript exists, HADEF transcribes via Whisper. This is where most automated tools fail. The most important things your guest has said are usually in podcast episodes, panel discussions, and conference talks that no one has bothered to transcribe. ### 3. Concept Agent Reads everything Research and Transcript surfaced. Looks for the conversation your guest has not had yet. The unexplored thread, the contradiction in their public record, the question someone almost asked but never did. Returns a proposed episode title, premise, and five interview questions with source-grounded hints. ### 4. Pitch Agent Writes the actual outreach. Two formats by default: a LinkedIn DM and an email, both in your show's voice. Includes a verified verbatim quote from the guest as anchor. ### 5. QA Agent Runs the output through a quality rubric. Verifies every quote against its source URL. Counts specificity anchors. Checks transcript coverage. Scores premise structure and host-voice fit. Returns a 0-100 quality score with line-by-line feedback. If the score is below 70, the agent flags what needs reworking before you see the pitch. ## What is different Three things separate HADEF from every other tool in the category. ### Transcripts, not summaries Most AI booking tools call an LLM and ask "tell me about this person." That returns a summary built from training data with a six-to-eighteen month cutoff. HADEF reads the actual primary sources, including audio that has never been transcribed. The pitch references things your guest said two weeks ago, not two years ago. ### Show Brain Your show has a voice. The way you open episodes, the kinds of questions you ask, the topics you avoid, the framing you use for guests. Show Brain learns this from every approval, edit, and rejection you make. After ten pitches it has a clearer picture of your editorial identity. After fifty, the pitches sound like you wrote them. ### Golden Thread Every pitch carries a verified verbatim quote from the guest. Not a paraphrase, not a summary. The actual sentence, tied to its source URL, time-stamped if it came from audio. If a quote cannot be verified, it does not ship. That single rule is the reason HADEF pitches do not get hosts caught hallucinating. ## Pricing 14-day free trial includes 3 pitches. No credit card required to start. ## Who HADEF is for - You host a B2B podcast under 100 episodes - You interview senior leaders such as CCOs, founders, executives, or authors - You currently spend an hour or more researching each guest - You want pitches that reference what the guest actually said, not generic flattery - You care about your reputation enough to verify before you send ## Who HADEF is not for - You run volume plays with cold templates and do not want to slow down - You want a contact database, not pitch generation - You want a marketplace that surfaces willing guests rather than targeting specific people - Your show has more than a hundred episodes and you have a workflow that already works - You are a PR agency running 100+ outreach campaigns a week. Look at Podseeker or Rephonic instead. ## How HADEF compares to other tools Different jobs, different tools. Honest cuts: - vs. Wobaka : Wobaka tracks contacts. HADEF writes the pitch. - vs. Streak : Streak lives in Gmail. HADEF reads transcripts. - vs. PodMatch : PodMatch is a marketplace. HADEF goes after specific named guests. - vs. Captivate : Captivate bundles a booking calendar with hosting. HADEF works with any platform. - vs. Deelo : Deelo runs the show end to end. HADEF specializes in outreach quality. - vs. Less Annoying CRM : LACRM is a flexible contact CRM. HADEF is built for one workflow. - vs. Podseeker, PodPitch, Rephonic : PR databases for high-volume agencies. HADEF is for individual hosts. - vs. Monday or Asana : General PM tools adapted as light CRMs. HADEF was built for the booking workflow specifically. - vs. HubSpot or Salesforce : Enterprise sales CRM with multi-show overhead. HADEF is single-host pricing. For the full guide with a decision quiz, see the comparison hub . ## FAQ ### What is HADEF? HADEF is a SaaS for B2B podcast hosts. It researches a potential guest, identifies the unexplored conversation angle, and writes a personalized outreach pitch in the host's voice. Every pitch is grounded in real transcripts from the guest's prior appearances and includes a verified verbatim quote. ### Who is HADEF built for? B2B podcast hosts with under 100 episodes who interview senior leaders such as CCOs, founders, executives, and authors. Three archetypes: theme shows like Future of X, founder-led shows where a founder interviews peers, and branded company podcasts. ### How does HADEF differ from a podcast guest CRM like Wobaka or Streak? Wobaka and Streak track contacts and pipelines. HADEF writes the pitch. Most hosts use a CRM for tracking and HADEF for the outreach generation itself. They solve different parts of the workflow. ### How does HADEF differ from a marketplace like PodMatch? PodMatch matches you with guests who have opted in to be matched. HADEF lets you target specific named guests who have not opted in. Different category. Use PodMatch when you want speed and willing participants. Use HADEF when you want a specific person. ### What does it mean that HADEF is transcript-grounded? Most AI pitching tools call an LLM and ask it to summarize a person. That returns information limited by the model's training cutoff. HADEF reads the actual primary sources, including audio that has never been transcribed, by running it through Whisper. The pitch references things your guest said two weeks ago, not two years ago. ### What is the Show Brain? Your show has a voice. The way you open episodes, the kinds of questions you ask, the topics you avoid. Show Brain learns this from every approval, edit, and rejection you make in HADEF. After about ten pitches it has a clear picture of your editorial identity. After fifty, the pitches sound like you wrote them. ### What is the 0-100 quality score? Every pitch gets scored against a quality rubric before you see it. The rubric checks: verified verbatim quote present, specificity anchor count, transcript coverage, host fit, and premise structure. If the score is below 70, the system flags what needs reworking. You always see the score and the breakdown. ### How long does it take to generate a pitch? Roughly 10 to 20 minutes per guest, depending on how much public material exists. You give HADEF a name and a LinkedIn URL. It pulls research, transcribes audio if needed, develops the concept, writes the pitch, runs QA, and returns the package. ### What happens if HADEF cannot find transcripts for a guest? The QA agent flags low transcript coverage and refuses to fabricate. You get a partial package with whatever was available, plus a note explaining what was thin. You can decide whether to send, edit, or skip. ### How is the 14-day free trial structured? 14 days, 3 pitches, no credit card required to start. The trial uses the same pipeline and quality scoring as paid plans. After the trial you choose Starter, Professional, or Elite. Ready to try it? 14 days, 3 pitches, no credit card. Terms Privacy Cookies HADEF, tan@hadef.io --- ## Source: Pricing (https://hadef.io/pricing) # One pipeline. Three scales. Every plan includes the full 5-agent pipeline: research, transcripts, concept, pitch, and QA. SoftwareApplication offers (canonical JSON-LD on homepage): - **Starter:** USD 49/month. Five pitches per month. One show. 1 GB storage. - **Professional:** USD 199/month. Twelve pitches per month with rollover (caps apply). Three shows. 10 GB storage. - **Elite:** USD 499/month. Twenty-five pitches per month with rollover (caps apply). Unlimited shows. 50 GB storage. MCP access. Offer descriptions from schema: - Starter: "5 pitches/month, 1 show, 1 GB storage" - Professional: "12 pitches/month with rollover, 3 shows, 10 GB storage" - Elite: "25 pitches/month with rollover, unlimited shows, 50 GB storage, MCP access" - Done For You: Per-result pricing via PriceSpecification; pay only when a pitch is accepted; targets agreed up front. Rollover: Professional rolls unused pitches up to a stored cap (20). Elite rolls unused pitches up to a stored cap (40). Yearly billing saves 20% versus monthly. Cancel any time. 14-day free trial: three pitches, no credit card required to start. Same pipeline and QA as paid plans.