Sedale Turbovsky

CEO. Founder @opengrants_io Strategic Funding Specialist. Professional Human Being. Follow for tweets on #grants, #govtech, #publicfunding and #startups. #ODCT1

[opengrants.io @opengrants_io]

Need to launch a grant-tracking tool for your community? 🚀 With the OpenGrants API, you can go from 'idea' to 'MVP' in a weekend. Access thousands of funding opportunities with a single integration.

Tell us what’s in your dev pipeline!
https://t.co/5271QUtXDS

[Sedale Turbovsky @STurbovsky]

The interesting question isn't "will AI replace this job."

It's: "does cheaper delivery expand the market, or shrink the team?"

Elastic demand (most underserved sectors) → more roles. Fixed demand → fewer. That's the entire macro story.

[Sedale Turbovsky @STurbovsky]

Framework for any "AI employee" pitch:

- Hours displaced per month (most teams underestimate by 30-50%)
- Loaded $/hr (≈1.5-2x base after payroll, benefits, tooling)
- Subscription cost

If payback isn't under 30 days, the pitch or the scope is wrong.

[Sedale Turbovsky @STurbovsky]

86% of AI roll-up investors now say margin uplift is the biggest value-creation lever. Multiple expansion sits at 46%.

The old model was financial engineering. The new one is operational redesign with capital attached.

Not investment advice.

[opengrants.io @opengrants_io]

One API for all your grant data needs. 🌍 We aggregate federal, state, and local funding opportunities so you don't have to.

Integrate OpenGrants for the most comprehensive data available. 🛠️

Start building: https://t.co/5271QUtXDS

[Sedale Turbovsky @STurbovsky]

GSA's draft AI procurement clause (552.239-7001) makes contractors responsible for any "Service Provider" in the stack — including third parties that aren't subcontractors.

For small AI vendors selling federal, supply-chain visibility just became a competitive moat.

[Sedale Turbovsky @STurbovsky]

This mood will pass. This sentence will pass. The hand that writes it, the room it is written in, the century it is written in — all passing. To remember this is not to despair. It is to stop gripping so tightly things that were never meant to stay.

[Sedale Turbovsky @STurbovsky]

The most expensive part of an AI agent isn't the tokens. In every services-business cost audit I've seen, tokens are 1-4% of total cost. The rest is human review-minutes.

If you're not measuring that, you're not pricing the work.

[Sedale Turbovsky @STurbovsky]

Gov is quietly becoming a serious AI buyer.

This week: HHS using AI to read 5 years of audits across 50 states. UAE moving 50% of gov services to agentic AI in 2 years.

Document-heavy, regulated, fragmented — the markets software ignored. Where I'd build.

Reshare if useful.

[Sedale Turbovsky @STurbovsky]

The quiet line in this year's AI economics: companies watching token bills pass the salaries they were meant to replace.

Bain has leaders modeling a future opex mix near 70–80% headcount, 20–30% tokens. Even at 1–2% today it strains budgets.

"Cheaper" isn't "cheap."

[Sedale Turbovsky @STurbovsky]

How to scope your first production agent without it blowing up:

Phase 1 (2–4 weeks): ONE bounded use case. Clear goal, measurable success, a real test set. Prove it works, set a baseline.

Don't expand scope until that passes. Most failed agents skipped this and started big.

[Sedale Turbovsky @STurbovsky]

The AI roll-up pitch rests on one assumption: AI delivers the margin gains operators historically didn't.

Yale's search-fund data: EBITDA went 25% → 19% entry to exit. Multiple expansion did the lifting, not operating skill.

Price the assumption, not the pro forma. Not advice.

[Sedale Turbovsky @STurbovsky]

I was not made to be self-sufficient in every way. The Stoic life is not a life apart but a life woven in — honest with friends, fair to strangers, patient with family. Virtue practiced alone is only half-practiced.

[Sedale Turbovsky @STurbovsky]

Operator note: picking which workflow to turn into an agent first.

Went with the boring one — weekly reporting. Repetitive, bounded, clear success criteria. That's where agents survive production; the flashy pick is the worst place to start.

Reply/DM if you're scoping yours.

[Sedale Turbovsky @STurbovsky]

We didn't lose 16k US jobs/month to AI. We lost the bottom rung.

A senior person with AI tools can do an intern's work. They can't do an intern's learning.

Operators with AI leverage now choose: keep training the next layer, or hollow out the management pipeline for 2031.

[Sedale Turbovsky @STurbovsky]

@Bencera this is amazing!

[Sedale Turbovsky @STurbovsky]

the future is here...

[Ben Cera @Bencera]

Polsia just raised $30M at a $250M valuation.

Approaching $10M annual run rate.

One Founder + AI. Zero employees.

Polsia runs companies autonomously.

It also ran its own fundraising.

I just showed up for signatures.

[opengrants.io @opengrants_io]

SBIR is back through 2031. The footnotes worth reading:

— Phase II "strategic breakthrough" track, up to $30M
— Foreign-affiliation due diligence on owners + key personnel
— Per-firm caps (NIH: 9 proposals/year)

Fewer shots. More weight on each. Pick well.

[Sedale Turbovsky @STurbovsky]

@Zecca_Lehn People building and owning small fleets of money making machines...."selling to your village".... community, reputation and distribution become the real moats...

[opengrants.io @opengrants_io]

NIH updated SBIR/STTR technical assistance (TABA) policy on April 13.

Phase I: up to $6,500 per project. Phase II: up to $50K.

Eligible uses: market validation, IP protections, cybersecurity, sales support.

Must be requested in the original application. #SBIR