Venture capital firms are deploying AI to accelerate fundamentally flawed processes, argues Henrik Landgren of Gilion. Rather than simply automating existing workflows, investors should prioritize building direct data connections to companies' financial, payment and accounting systems. This approach would transform due diligence by accessing raw operational data instead of founder-packaged presentations.

Landgren, former VP of analytics at Spotify, witnessed firsthand how granular data revolutionized decision-making during his tenure at the music streaming giant. He contends that today's AI adoption in investing focuses on generating reports faster rather than eliminating unnecessary processes. The core issue remains garbage in, garbage out – LLMs perform only as well as their underlying data quality.

Direct data integration shifts investor risk assessment dramatically. Instead of starting from zero with cherry-picked information, analysts could begin at 70% completion, reserving human judgment for evaluating team dynamics and founder potential. This methodology particularly benefits overlooked startups that lack flashy presentations but demonstrate solid fundamentals.

The market implications are significant. Capital-efficient businesses with strong retention often get passed over because their compelling data isn't accessed quickly enough. Better infrastructure gives investors a competitive edge in deal speed while enabling more confident funding decisions for emerging sectors like AI-powered hardware and deep tech that traditional metrics cannot evaluate accurately.