Griffin Krum

This website has DR 80. Why are you rejecting it?"
That was my client's question.
My answer surprised them:
"Because only 20% of its content is actually about SaaS."
That one decision completely changed our link-building strategy.
A client recently sent me a list of high-DR websites for a SaaS backlink campaign.
The first question wasn't about price.It was:
"Are these actually SaaS websites?" At first glance, they looked perfect.High DR. Strong traffic. Affordable pricing. But when I reviewed the content, I found a problem. Most of the sites only published an occasional SaaS article.
The rest of their content was about finance, lifestyle, travel, gaming, or general tech. For a SaaS company, that's not enough.Instead, we focused on publishers where 70–80% of the content was consistently about topics like:
👉 AI & Generative AI
👉CRM & Sales Software
👉 Marketing Automation
👉Cybersecurity
👉Cloud Computing
👉 DevOps & APIs
👉 Productivity & Collaboration Tools
👉 B2B SaaS & Enterprise Software
Yes, the pricing was slightly higher. But the relevance was dramatically better.
A backlink isn't just about DR anymore. It's about whether your brand belongs in the conversation your audience and increasingly AI associates with your niche.
After replacing general tech placements with highly relevant SaaS publishers, the client's backlink profile became far more aligned with their business, creating stronger topical authority instead of just increasing the number of links.
Lesson: Stop negotiating only on price.
Start evaluating the percentage of SaaS relevance on every publisher you buy from.Because in 2026, context beats cost, and relevance beats raw metrics.

NOW HIRING (apparently): Full-Time Email Re-Typer. No experience required. You already have the job.

Let me read you the description, because you applied without noticing.

ROLE: Manually retype the same ~30 messages every working day. The thank-you. The intro. The "just circling back." The "sounds good, let's set it up." Type each one from scratch, every time, as if you've never seen it before.

HOURS: ~4 per week, indefinitely. Roughly 11 working days a year.

COMPENSATION: Paid entirely in exhaustion. Bonus structure: a vague sense at 6 PM that you were busy all day but can't say doing what.

REQUIRED SKILLS: The ability to ignore that a computer could do this instantly. A high tolerance for typing "hope you're doing well" while feeling nothing.

CAREER GROWTH: None. The job is identical in year five.

—

Here's the thing. You're overqualified for this role. You were hired for something else — the actual work, the thinking, the stuff only you can do. But this little copy-typist job got bolted on so quietly that you forgot it's optional.

So consider this your two-weeks' notice. To yourself.

The replacement is free, starts in 60 seconds, and never gets tired. You write each message once, turn it into a shortcut, and from then on it types itself — a few letters, the whole thing appears. When a new kind of message comes up, you describe it and AI drafts it.

You don't lose the job. You get promoted — back to the work you were actually hired to do.

The Re-Typer position is now vacant. Leave it that way.

→ keytext.app (free)

What's the most absurd task you do by hand that a machine should've taken over years ago? Let's write the job posting for it together.

#Productivity #FutureOfWork #Automation #CareerGrowth #WorkSmarter

🚀 HRPGD

Great organizations are built by great people.

Success starts with hiring the right talent, developing strong teams, and creating a workplace where people can thrive. When people grow, businesses grow with them.

👥 Empower talent

📈 Build high-performing teams

💡 Invest in continuous growth

🌍 Shape the future of work

The strongest competitive advantage is a team that never stops learning.

HRPGD — Empowering people. Accelerating growth.

#HRPGD #HumanResources #Leadership #TalentDevelopment #WorkplaceCulture #BusinessGrowth #ProfessionalDevelopment #FutureOfWork #PeopleFirst #Success

I built PatchWork because every AI resume tool I tried still made me do the work.

They'd suggest bullet points or rephrase a sentence. Cool — but I'd still spend 45 minutes per application customizing my resume and writing a cover letter from scratch.

PatchWork works differently. You upload your career history once — past resumes, LinkedIn, whatever you've got. Then you paste a job description, and it writes you a fully tailored, ATS-ready resume and matching cover letter in about 2 minutes. Not suggestions. Not edits. The whole thing.

Early users are getting interviews on ~7 out of 10 applications.

The pricing model was a deliberate choice too. No subscriptions. You buy credits — 1 credit = 1 job application. Your first one is $2.95, and a 20-pack is $25 ($1.25/ea). Credits never expire. Job seekers are already stressed about money; a recurring charge while you're between jobs felt wrong.

Bootstrapped, profitable unit economics, and growing through Google Ads + word of mouth.

Would love any feedback: usepatch.work

Excited to share this after 3 months of relentless building and shipping 🥳

Introducing PrepAlly - an AI mock interviewer that tries to replicate the real experience instead of just testing questions.

Why did I build PrepAlly?
Interviews are strange. You can spend weeks preparing, revise frameworks, rehearse answers in your head... and then the interviewer asks one unexpected follow-up and suddenly everything falls apart.

The truth is: most people don’t struggle because they lack skills. They struggle because they’ve never practiced interviewing in a realistic way.

That’s exactly why I built PrepAlly. It is designed to help you practice real interviews, anytime, anywhere.

Here’s what makes it different 👇
- 🎙️ Live AI voice interviews personalized to your resume, experience, and target role
- 🔗 Generate interviews from a job link - paste a job posting and PrepAlly builds a tailored interview around the role requirements
- 🧠 Dynamic interviewer personas that adapt to the job description - so a PM interview feels different from consulting, tech, finance, or MBA interviews
- 📄 AI-powered resume analysis to identify strengths, gaps, and areas you should be ready to defend in interviews
- 📈 Detailed feedback beyond just answers - confidence, communication, tone, delivery, vocabulary, and even posture/body-language analysis to help you improve how you present yourself, not just what you say

and more...

Whether you're preparing for internships, career switches, or aiming for top tech, consulting, finance, or leadership roles - PrepAlly helps you practice until confidence feels natural.

Try it free at prepally.app

Traditional LMS platforms take forever set up and require dedicated teams to run, making them impractical to start-ups or companies without a dedicated L&D department. I intend to fix that with Mentorra; an onboarding platform that takes ten minutes to set up. You drop in your training materials, create roles, invite employees, and Mentorra automatically organizes your messy content into structured, bitesized modules. It's also got an AI assistant that can help answer common trainee questions so trainers don't have to repeat themselves over and over again.

The hope is to make onboarding less of a headache for start-ups and small companies.

I'll attach a link to the demo below. I'd really appreciate it if you guys could take a look and answer a few questions;

- does the value proposition make sense?

- what's missing?

- would you use this?

Thank you!

mentorra.lovable.app/

We are witnessing a bizarre double standard in enterprise AI.

​When a company hires a fresh employee, they don’t hand them a laptop on day one and expect them to magically know the entire corporate strategy, past client wins, and internal pricing rules without training. They onboard them. They give them an internal handbook. They teach them.

​Yet, executives are firing experienced humans, replacing them with a raw LLM API, and then getting angry when the agent hallucinates or fails.

​AI isn't magic it’s trained on human experience. If you don't give your agents a dedicated, structured "Company Brain" to learn your specific business context, you haven't hired an automated workforce. You’ve just hired an army of interns with amnesia. Stop trying to automate your workflows until you build the infrastructure to actually train your agents.

I think I'm about 2 weeks out from shipping the Job Seeker portal of the TRAWLR product, giving candidates a better search, application management and preparatory tools for landing their next gig.

Once she's up and running I can fully focus on the Recruiter portal 🥂

Oddit is a dev evaluation engine that turns a candidate's real shipped code into a 5-minute hire memo with file-and-line evidence and 11 work-pattern signals on how they actually engineer.

Hiring developers in 2026 has broken in a specific way. Candidates use AI to write their resumes. Engineers use AI to write the code they ship. Greenhouse's 2026 AI in Hiring Report (n=4,136) found 86% of recruiters caught or suspected candidate fraud last year, and 42% of candidates admit prompt-injecting their resumes. Meanwhile JetBrains' January 2026 survey of 10,000+ developers shows 90% use AI tools at work to write code.

So hiring teams have never had more evidence about a candidate, and never had less signal about whether the work is actually theirs or how they actually think. Bersin pegs the all-in cost of one bad senior engineering hire at $375k to $500k. Hiring managers are spending an hour or more reading each candidate's code and projects before a final-round call, guessing at what's AI-generated and what isn't, and still getting it wrong.

We solve this by auditing the candidate's actual code and producing a hire memo in 5 minutes. The memo classifies what they wrote (custom vs framework glue vs AI-assisted), cites file-and-line evidence for every claim, and surfaces 11 work-pattern signals (rationale habit, iteration discipline, decision quality, scope management, and 7 more) so the hiring team sees not just what they shipped but how they actually engineer.

Built and shipped the engine. Working multi-model audit pipeline in production: Code Property Graph analysis across 15+ languages via tree-sitter, 17 deterministic pattern detectors, a tool-calling verification agent that re-checks every claim against the actual code, and 11 person-level work-pattern signals across all of a developer's audited work. 153 audits run on real developer code since the May launch. 77 users signed up.

weoddit.com

most founders build for months before finding out nobody wanted it.

I almost did the same thing.

before I wrote a single line of code for my current project, I forced myself to answer one question: is there actually demand for this?

turns out that question is harder to answer than it sounds. surveys are biased. asking friends is useless. hiring a research firm costs more than your entire runway.

so I built a way to simulate it instead.

when a complete stranger — someone I've never met, never marketed to — signed up on their own, I knew I was onto something real.

that's the only validation that matters. not compliments. not 'looks cool bro'. a stranger taking action.

if you're pre-launch and still guessing whether your idea has legs — ScaleSim can tell you before you waste months finding out the hard way.

AI-powered market validation in one clean and simple report.

→ scalesim-waitlist.netlify.app

pgblame — CI guards your code and ignores your database. pgblame watches
Postgres query performance across deploys, so you see which deploy slowed a
query. $19/mo, real free tier. Looking for 2 design-partner beta users.

Backstory: I've had plenty of Postgres pain over the years, but the thing
that always nagged me wasn't one incident — it's that CI guards your code and
goes silent on your database. Deploys ship; nobody's watching whether they
slowed a query.

So I built the tool I wanted: pgblame. A tiny read-only Docker agent
snapshots pg_stat_statements every 60s, ingests your Vercel/Railway/GitHub-
Actions deploy webhooks, and shows "this query went 40ms → 800ms right after
this deploy." Same shape as Lantern but not Rails-only; ~1/8 the price of
pganalyze ($19/mo, real free tier). The agent runs in your env, read-only,
never sees your data, MIT-licensed.

It's live at pgblame.com but I haven't launched publicly yet — I want
2 founders running Postgres in prod (Supabase/Neon/RDS + Vercel/Railway,
shipping often) as design partners: 20-min call, you set it up while I watch,
brutally honest feedback, free Pro forever. Who's in?

Job boards waste hours on bad JD data. Here's why — and how to fix it in one API call.

When an employer posts a job, they paste raw text. Your database gets a blob. No structured salary. No normalised skills. No seniority tag. Filtering and search end up useless.

I built JD Intelligence to fix this. One API call returns 16 structured fields from any raw job description: title, salary range, required skills, seniority level, remote policy, visa sponsorship, and more.

Also: /score (resume-to-JD matching), /enrich (rewrites JDs for clarity), /classify (department + seniority tagging).

$29/mo. 7-day free trial. Instant API key. No sales call.

Live demo — no signup: jdintelligence.dev/demo

Built for job boards, ATS tools, and recruiting agents. If you're building in this space I'd love to hear what you're working on.

Hello everyone, I built a job description API that can parse, score, enrich, and classify job descriptions with a single API call. Affinda charges $800/mo for JD parsing with a sales call requirement. I built the self-serve alternative starting at $29/mo with a 7-day free trial for a limited time.

Still in development, so there's more to come.

Demo page on the site, let me know your thoughts.
JD Intelligence — Job Description API