Software-driven diligence, dedicated GP-led secondaries, and sharper investor execution
The gist
VC work is shifting from manual review and ad hoc deal structuring to criteria-encoded diligence and specialized continuation-capital execution.
This week’s developments
Arch Brings Criteria-Driven Screening Into the Diligence Layer
Arch launched an AI-powered Investment Research tool for pre-investment diligence that ingests offering materials, extracts key terms, links outputs back to source documents, and tests deals against a firm’s own investment criteria. Arch says early users can cut diligence time by up to 50% per deal. The important shift is not generic document summarization; it is embedding a fund’s screening rules into the review layer while preserving auditability through source-linked outputs.
That extends the workflow shift from last week: once technical diligence becomes more repeatable, the next bottleneck is first-pass screening, and Arch is pushing that step into software as well. The result is a narrower but more actionable change in VC workflows: first-pass diligence is becoming criteria-driven and software-mediated, even if underwriting itself is not yet standardized. For investors and platform teams, the advantage will accrue to firms that can define their investment tests crisply and operationalize them consistently. Teams still doing this manually will spend more time on repetitive screening work that is increasingly being productized, while firms that adopt tools like this can move faster from inbound materials to a yes/no decision.
How should we codify our screening criteria for AI diligence?
If you're an individual contributor
- Manual first-pass diligence is shrinking; criteria review is the new edge.
- Get sharp at checking AI outputs against firm criteria and source docs; that’s how you stay useful as screening gets automated.
Sources
- Run your pitch deck through a VC’s AI screener before they do. — Venture Curator, September 17, 2026
Use a VC-style AI rubric to test decks against criteria like traction, LOIs, and why-now before pitching.
- Once AI Reads the Deck, Venture Investors Test What the Data Cannot Explain - KoreaTechDesk | Korean Startup and Technology News — KoreaTechDesk, October 5, 2026
Shows where AI diligence helps, where humans still catch risk, and how to combine both in VC workflows.
- 3 Founders Shipped Claude Managed Agents to Production in 2 Weeks. Here's the Playbook — The AI Corner, September 24, 2026
A playbook for dogfooding, external eval sets, and guardrails to improve live AI systems.
If you manage a team
- Your team’s bottleneck is shifting from reading decks to judging exceptions.
- Coach analysts to define criteria crisply and spot edge cases fast; repetitive screening is becoming software work.
Sources
- Microsoft releases new AI playbook for enterprises with real-world examples, and it reveals a surprising 'moat' you may already have — VentureBeat, September 17, 2026
Framework for mapping processes, defining human-agent boundaries, and using proprietary evals and context to speed decisions.
- The AI-native SDLC won't be one process — The New Stack, September 12, 2026
Shows how to route work by risk, preserve audit trails, and reserve human judgment for edge cases.
- Why AI Adaptation, Not Adoption, Is the Real Work Ahead — AI Marketing Institute, September 21, 2026
Framework for redesigning processes, coaching teams, and building repeatable AI-assisted work habits.
If you lead the organization
- Firms that codify investment tests will outpace those still screening by hand.
- Push for criteria-driven diligence workflows and source-linked auditability; manual first-pass review is now a scaling constraint.
Sources
- Episode 25: Human in the Loop Is Not a Strategy — Finding 12 Minutes Podcast, September 22, 2026
Framework for defining decision rights, review criteria, and metrics in AI-enabled workflows.
- Where AI Should Actually Earn Its Keep in Biopharma - with Mukhtar Ahmed of Greenstone Biosciences — The AI in Business Podcast, September 29, 2026
Framework for applying AI to explicit decision criteria, evidence synthesis, and portfolio-level diligence quality.
- The Prudent AI Series — Legal Tech Monitor, October 2, 2026
Framework for setting kill criteria, preserving optionality, and making AI investments through fast, reversible experiments.
New Mountain Builds a Dedicated GP-Led Secondaries Platform
New Mountain launched New Mountain Atlas I, a dedicated GP-led secondaries fund targeting up to $2 billion and focused on single-asset continuation transactions. The fund is designed to provide lead or anchor capital for “quality companies” that cannot find a traditional exit, and New Mountain says it has been building the platform since early 2025. That moves continuation deals from an occasional structure into a formal capability.
The launch fits New Mountain’s prior activity, including its role in the Corsair CV and progress on a multi-asset continuation vehicle for Real Chemistry. Together, those moves show a manager building a repeatable GP-led platform around mature sponsor-owned assets, not just opportunistic one-offs.
For venture and growth teams, the progression is clear: continuation transactions are becoming a more standard liquidity route for late-stage companies that are not ready for a sale or IPO. That makes valuation discipline, governance, and asset quality even more important in every process you run or join.
How should we adapt our exit strategy for continuation deals?
If you're an individual contributor
- Continuation deals are now a real exit path, not a side show.
- Sharpen diligence on governance, valuation, and asset quality; those skills will matter more than chasing the fastest sale.
Sources
- I Ran a Startup Through a VC Investment Committee. It Died in Four Minutes. — The Founders Corner®, September 21, 2026
Step-by-step guidance for presenting traction, risks, and economics without getting disqualified by inconsistent inputs.
- Is the SaaS crash narrative wrong? Here’s what 72,000 companies show. | Claude is already leading 26% of Anthropic’s AI R&D. — Venture Curator, September 18, 2026
Shows how to use simple models, burn metrics, and assumptions to assess founder quality and business health.
- Buy-side readiness: Protecting value before the deal is done — Africa Private Equity News, September 24, 2026
A practical framework for diligence, governance, and integration planning to preserve value before and after closing.
If you manage a team
- Your team needs to judge hold-vs-exit quality, not just run processes.
- Coach analysts and associates to pressure-test continuation options, sponsor incentives, and downside risk in every late-stage process.
Sources
- The Exit Playbook: Right-timing the exit — Private Equity Spotlight, September 29, 2026
Framework for timing exits, building management strength, and preparing clean diligence materials to maximize value.
If you lead the organization
- GP-led secondaries are becoming a core liquidity lane for mature assets.
- Rework your late-stage playbook: build continuation expertise, tighten valuation standards, and decide where your firm can lead.
Sources
- The Continuation Fund Boom Comes With Growing Regulatory Questions — Institutional Investor Knowledge Center, September 29, 2026
Explains SEC scrutiny, conflict management, valuation disclosure, and fairness-opinion practices for continuation deals.
- Well-Caffeinated: Not Just a Speed Bump — Identifying and Managing Conflicts of Interest in Private Funds — The Freshfields Podcast, August 26, 2026
Explains valuation, disclosure, and LPAC processes for handling conflicts in GP-led secondaries transactions.
- Golub Capital’s David Golub - inside the mind of one of private credit’s pioneers — Alt Goes Mainstream: The Latest on Alternative Investments, WealthTech, & Private Markets, September 10, 2026
David Golub explains how debt-market strengths can translate into a scalable GP-led secondaries business.