Tim Tully, a partner at Menlo Ventures, is leveraging his deep technical background as a former programmer and CTO to build an impressive portfolio of AI startups. The venture capital firm, historically known for backing consumer giants like Chime, Warby Parker, Roku, and Uber, has pivoted significantly toward generative AI investments.
Tully’s journey began at age 6 when he taught himself to code on an Apple II computer, creating text-based adventure games using Applesoft BASIC. His technical expertise spans decades, including 14 years at Yahoo where he led engineering for Yahoo, AOL, and media properties like Huffington Post and Tumblr, managing over 850 people. He later served as CTO at Splunk, overseeing the company’s critical migration to the cloud before joining Menlo Ventures as a partner in 2021.
Menlo Ventures has made a substantial commitment to AI, closing a $1.35 billion fund specifically for investing in early-stage AI startups. The firm has become an early backer of several AI powerhouses, including Pinecone (vector database), Anthropic (LLM developer), Unstructured (enterprise data cleaner), Neon (conversational AI), and Cleanlab (automated data cleanup).
Since joining Menlo, Tully has invested in nine startups, including Pinecone’s $28 million Series A (2022), Neon’s $46 million Series B (2023), Unstructured’s $40 million Series B (2024), and JuliaHub’s $24 million raise (2021). His portfolio company TruEra, an AI quality-management software startup, was acquired by Snowflake earlier this year.
What sets Tully apart is his hands-on approach: he maintains a data center in his basement, regularly deploys software, tests startups’ APIs, and even writes code for portfolio companies. In one notable instance, he wrote a software-development kit for fintech startup Orb, helping secure a $14 million Series A led by his colleague Naomi Pilosof Ionita.
Tully is particularly excited about opportunities at the application layer, focusing on vertical AI categories including legaltech (he wrote a $14 million seed check for AI legal assistant Eve in 2021), marketing, and automation. He personally uses AI coding copilots, reporting they make him “50% to 60% more productive.” His wife, Jess Lee, is a partner at Sequoia Capital, further connecting him to the venture capital ecosystem.
Key Quotes
I’m still coding every day. That really helps me for my job, and I think it’s really hard to be a VC and invest in a deeply technical product if you don’t have a technical background. I’m actually using this stuff.
Tim Tully explains his hands-on approach to venture capital, emphasizing how his daily coding practice gives him a competitive advantage in evaluating AI startups. This distinguishes him from traditional VCs who rely primarily on market analysis rather than technical validation.
My background helps me connect with technical founders because they want investors who they can talk shop with. I’ve managed teams and have been CTO of a public company.
Tully describes his winning strategy for building relationships with AI startup founders. His operating experience as Splunk’s CTO and managing 850+ people at Yahoo gives him credibility that purely financial investors lack.
I’m really high on AI coding assistants. Opportunities in this space are pretty compelling.
Tully shares his enthusiasm for AI coding copilots after personally experiencing 50-60% productivity gains. This firsthand validation from a technical investor signals strong potential for AI developer tools as an investment category.
Our Take
Tully represents a new breed of venture capitalist who bridges the gap between technical expertise and investment acumen. His approach—maintaining a home data center, testing APIs, and even writing code for portfolio companies—demonstrates that successful AI investing requires more than pattern recognition and financial modeling. It demands genuine technical understanding.
The $1.35 billion AI fund at Menlo Ventures reflects broader market dynamics: institutional capital is flooding into AI, but increasingly flowing through investors who can technically validate claims. Tully’s portfolio strategy is particularly astute, balancing infrastructure plays (Pinecone, Anthropic) with application-layer bets (Eve, Orb) across multiple verticals. His personal productivity gains from AI coding assistants aren’t just anecdotal—they’re market validation from someone who actually understands the technology. As AI moves from research labs to production environments, investors like Tully who can evaluate technical merit will likely outperform those relying solely on market narratives.
Why This Matters
This story highlights a critical trend in venture capital: the increasing importance of technical expertise in evaluating AI investments. As generative AI transforms from hype to practical applications, investors with hands-on technical backgrounds like Tully have a significant advantage in identifying promising startups and avoiding overhyped technologies.
Menlo Ventures’ $1.35 billion AI-focused fund signals major institutional capital flowing into the AI infrastructure and application layers, validating the sector’s long-term potential beyond just large language models. The firm’s portfolio strategy—spanning vector databases, data cleaning, conversational AI, and vertical applications—demonstrates how the AI ecosystem is maturing beyond foundational models into specialized tools and industry-specific solutions.
Tully’s success with companies like Pinecone, Anthropic, and the Snowflake-acquired TruEra shows that early-stage AI infrastructure investments can deliver significant returns. His focus on application-layer opportunities, particularly in legaltech and coding assistants, suggests where the next wave of AI value creation may occur. For founders, this underscores the importance of engaging with technically sophisticated investors who can provide strategic guidance beyond capital.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources: