Ascend (Ascend.vc)

Firm Profile

Seattle’s most active pre-seed firm — prefers to be the company’s first institutional check in the $250K–$750K range (sweet spot ~$500K), ~35–40 investments per fund (GeekWire + f4.fund + Ascend “about” all agree). Founding General Partner Kirby Winfield launched the firm in 2019 after two decades leading four startups. Thesis (Winfield’s own framing): “SaaS 3.0” (vertical AI agents/applications), “Gen AI 2.0” (horizontal AI/infra fast-followers), and Frontier AI (AI applied to the physical world’s hardest problems).

Funds / vintage:

• Fund I — $15M (2019)
• Fund II — $25M (2023)
• Fund III — raising as of 2025 (SEC filing surfaced by GeekWire; size undisclosed at filing). ^[single-source on “currently raising”]

Geography: Seattle-anchored but invests beyond the PNW. Structure: small partnership; Winfield is the lead/face, Nate Bek co-authors most deal memos.

Partners

• Kirby Winfield — Founding General Partner. Ex-founder (“recovering founder”); 20 yrs across four startups before VC. Sets the thesis and leads deals. (No init.inc relationship — no partner stub created; this firm is DEAD-tier.)
• Nate Bek — co-authors Ascend investment posts (associate/principal-level). ^[inferred from blog bylines]

Recent pre-seed/seed activity (Pass-1 sketch — ≤5 deals)

Ascend’s own blog is the cleanest source; it announces nearly every check. Pattern: participates in (rather than leads) most rounds, joining a stronger lead. Recent:

• AZX — $6M pre-seed, Jan 2026 (AI2 Incubator led; Ascend participated). Vertical AI for critical infrastructure.
• Govstream.ai — $3.6M seed, Dec 2025 (47th Street Partners led). AI permitting for local gov.
• Hyphen AI — $5M seed, Oct 2025 (Unlock Ventures led).
• Hearvana — $6M seed, Nov 2025 (Point72 led). ^[single-source on amount]
• Oumi — $10M seed, Jan 2025 (Venrock led). Open-source AI platform.

Full data → /Users/sazzad14/seed-market-scan/firms/ascend.csv.

Best data channels for this firm

Ascend’s own blog (ascend.vc/blog) is the single best source — Winfield/Bek post a deal memo for nearly every investment with round size + lead named, far more complete than press. GeekWire (Seattle beat) is best for fund-level news (raises, SEC filings). Crunchbase/PitchBook understate Ascend’s role (it’s usually a non-lead participant, so press credits the lead). Signal.nfx good for Winfield’s personal deal history.

Fundraise Relationship (init.inc)

• Status — DEAD (no viable path): Thesis fit was plausible (Ascend’s “SaaS 3.0 / vertical AI agents” thesis overlaps init.inc’s applied-AI-lab-for-back-office-work positioning, and Ascend leans pre-seed where init sits). But the only connector was Isaiah, who is now joining init.inc as a co-founder and can therefore no longer act as a neutral intro source. No alternate warm path was identified. Stage/check size (~$500K)issmallrelativetoini{t}^{\prime }s$7M target — Ascend would at most have been a participant, not a lead.
• Conversation log (newest first):
  • 2026-04-28 — Path LOST / closed. “No remaining path to firm. Isaiah was the only connector here, and he is now joining the team as a co-founder so can no longer act as an intro source.” (Same closure batch as First Round, GreatPoint, South Park Commons, Madrona — all Isaiah-only paths.)
• POC: Kirby Winfield (Founding GP) — would have been the target, but no intro was ever made. Intro source: Isaiah (now co-founder, conflicted out).
• Mutual connections / warm path: None remaining. Isaiah (init co-founder, plain-text per link discipline) was the sole connector and is now conflicted out as an intro source.
• Competitor-conflict flag: None found. No Ascend position in any of init.inc’s 8 tracked AI-ITSM competitors (Serval, Console, Atomicwork, Edra, STLabs, Treeline, Risotto, Ravenna) per cap-table-patterns-across-startup-competitors. Ascend would be a clean lane if a path ever reopened — the kill is purely connector-loss, not conflict or thesis.
• Personal & rapport notes: Winfield is publicly very online (active LinkedIn, Delphi AI clone, blunt “GTM / get this money” content) — a cold/semi-warm approach is plausible if init ever wants to reopen, given how accessible he is.

Cross-check vs existing wiki

No prior mention of Ascend in cap-table-patterns-across-startup-competitors or any competitor cap table. No contradictions. This is a net-new clean-lane firm killed only on connector availability.

Related

niam (sibling Isaiah-connector-loss closures — same 2026-04-28 batch) · cap-table-patterns-across-startup-competitors

Pass 2 deal pattern (2026-05-27)

Verified (all 5 Pass-1 rows carried forward unchanged):

• ✅ AZX ($6M pre-seed, Jan 28 2026, AI2 Incubator led, Ascend participated) — confirmed via Ascend blog + Exa.
• ✅ Govstream.ai ($3.6M seed, Dec 4 2025, 47th Street led, Ascend participated) — confirmed.
• ✅ Hyphen AI ($5M seed, Oct 26 2025, Unlock Ventures led, Ascend participated) — confirmed.
• ⚠️ Hearvana ($6M seed, Nov 5 2025, Point72 led, Ascend participated) — confirmed Ascend participation; amount still single-source per Pass-1 note.
• ✅ Oumi ($10M seed, Jan 29 2025, Venrock led, Ascend participated) — confirmed.

Net-new in Pass 2 (1 deal):

• ✅ Limelight — ~$2M seed, ~Feb 2026, SNR Capital led, Ascend participated. AI platform automating personality-led growth for B2B (B2B influencer marketplace + creator-marketing OS). Founder David Walsh (3x founder, 2 exits, ~40K LinkedIn followers); CTO Youngjae Ji (ex-Uber-endorsed). Confirmed via Ascend blog post + Ascend’s own “LinkedIn is the new enterprise media” essay (Feb 18 2026, which discloses Ascend’s Limelight investment) + Ignite podcast episode with David Walsh (Apr 2026). Date inferred from Ascend blog-post date.

Lead-rate / check / stage pattern: Pass-2 fully confirms Pass-1’s read — Ascend is almost always a non-lead participant in $2M-$10M pre-seed/seed rounds where a stronger lead (AI2 Incubator, Point72, Venrock, 47th Street, Unlock, SNR) anchors the round. Ascend’s $250K-$750K check fits this participation model. The firm’s own blog (ascend.vc/blog) remains the single best data channel — every recent deal landed there with round size + lead named, and Pass-2 surfaced one net-new (Limelight) from the same source plus the firm’s adjacent “LinkedIn is enterprise media” essay. Best data channel (Pass-2-confirmed): Ascend’s own blog is canonical and complete; GeekWire for fund-level news (Fund III raise filing); no need to scrape regional press or aggregators. The classification of Ascend as DEAD on connector-loss (Isaiah now init co-founder) stands — and Ascend would remain a small-check participant even if reopened (sub-scale for init’s ~$7M target).