Why Blue Lakes exists
Most teams have AI tools. Few see meaningful ROI. The gap isn't better models or more features — it's infrastructure.
Individual AI tools like ChatGPT are powerful for individuals. But knowledge work happens collaboratively. Teams need infrastructure that enables conversational work — where multiple people and AI work together, connected to organizational knowledge, with full provenance on every response.
That's the infrastructure Blue Lakes provides.
This shift requires infrastructure, not just better AI models. Blue Lakes provides that infrastructure.
The 86/14 paradox
Why? Most teams use AI the way they used Google in 2005 — individually, for basic search. That's consumer-level use.
Real business value requires transformation:
- —Teams working conversationally, not individuals using AI in isolation
- —AI connected to organizational knowledge, not generic responses
- —Provenance showing where information came from, not unexplained outputs
Individual tools like ChatGPT Teams don't enable this transformation. They give everyone individual AI access and call it “team collaboration.” But conversations stay siloed, organizational knowledge stays disconnected, and sources stay hidden. Blue Lakes provides the infrastructure teams need to progress beyond consumer-level use toward real organizational capability.
Built for teams, not individuals
Blue Lakes is explicitly for teams and knowledge workers, not individuals.
If you're an individual looking for AI assistance, ChatGPT or Claude are excellent tools. We're not competing with them for individual use.
But if you're a team that needs to:
- —Collaborate with AI, not just use it individually
- —Access organizational knowledge through AI
- —Verify sources and trust AI outputs
- —Demonstrate ROI from AI investments
Then individual tools can't get you there. They weren't designed for team-level transformation. Blue Lakes was.
Collaborative workspace
Multiple people and AI work together in the same conversation. Not individual sessions that stay siloed like ChatGPT Teams, where everyone remains in their own bubble.
Connected organizational knowledge
AI accesses your docs, repos, and past work, and keeps it connected across all your team's conversations. Not generic AI that knows nothing about your organization and forgets everything after each session.
Work that builds on itself
Every conversation, project, and decision becomes part of your organization's growing knowledge base. Your team gets smarter over time, not just faster in the moment.
Guide adoption, don't micromanage
See how your team uses AI and what's working, without surveilling individual sessions. Drive adoption and quality across teams, not just individual productivity.
Plus: vendor independence (works with Claude, GPT, any AI model), customer ownership (vault-native architecture keeps your data in your environment), and full provenance on every response (sources cited so you know where answers came from, not unexplained AI outputs).
Who's building Blue Lakes
Chevan Nanayakkara
Founder
Chevan brings 20 years of experience spanning datacom/telecom, market research, data monetization, adtech, and media, from single-digit-employee startups to global corporations. He has held executive and staff roles across product, operations, and implementation, giving him a practitioner's view of how organizations actually adopt new technology. Blue Lakes is where all of that converges: helping teams build sturdier bridges with their customers and their own knowledge by making AI a real part of how work gets done.
Founding Credibility: Built from Lived Experience
The Blue Lakes approach comes from lived experience building and operating conversational work infrastructure, not from abstract product research. Before building the product, Chevan built and operated the core patterns manually: conversation serialization, centralized knowledge storage, cross-repository linking, archive management, and configuration-as-code for AI behavior. These patterns emerged from solving real workflow pain across multiple projects and repositories.
This is the same pattern that produced Git: Linus Torvalds built it in 10 days because he understood collaborative development workflows deeply enough that the solution was almost obvious once he started. Deep workflow understanding, not technical wizardry, is the foundation of durable product design. The manual overhead of operating this system defines the exact product opportunity — the gap between “this works manually” and “this is easy for non-technical teams at organizational scale” is where Blue Lakes sits.
The Team: Execution Track Record
Blue Lakes is backed by the Invadraw team, a group that has shipped together before, reducing the execution risk that comes with first-time teams.
Enterprise infrastructure and data sovereignty expertise. Brings the architecture rigor needed for organizations that take data ownership seriously.
Architectural thinking and data rigor from large-scale scientific computing. The kind of analytical depth that shapes durable system design.
The delivery mechanism for the 90-day pilot model. Ensures every engagement produces real outcomes, not just access to software.
Connect
Where we're headed
Build your knowledge foundation
The 90-day pilot structures your team's knowledge so AI can actually use it — connected, organized, with full attribution. This is the foundation everything else depends on.
AI that builds on what your organization knows
With knowledge infrastructure in place, your AI doesn't start from scratch in every session. It draws on your team's past work, decisions, and expertise. It gets more useful over time, not less.
Ready for AI Agents
AI Agents are coming — purpose-built software that takes actions, makes decisions, and works alongside your team. They need reliable, organized knowledge to act on. Organizations that have built their knowledge infrastructure will be ready for them. Those that haven't will scramble to catch up.
The 90-day pilot is step one. If you're interested in being part of it, request a demo or reach out.