A recap of the on-demand webinar with Forrester, Cornell Lab of Ornithology, and LevelShift
| Quick Summary
• Cornell Lab of Ornithology manages roughly 8 petabytes of scientific and public-contributed data, including 1,000+ years of continuous bird-song recordings and millions of citizen-science contributions through eBird, Merlin, and Birds of the World.
• In an on-demand webinar with Forrester VP & Principal Analyst Charles Betz, Cornell Lab CTO Ananda Paramasivam and LevelShift’s Selva Pandian explain how they’re unifying that fragmented data into a single “Constituent 360” view across five business units.
• The core lesson: AI is only as good as the data foundation underneath it. Cornell Lab already runs mature AI on its science and mission platforms (like Merlin’s bird-ID models) — this project brings that same rigor to donor and member engagement before layering on more AI.
• Forrester recently recognized LevelShift alongside Accenture and Deloitte for its Salesforce Data Cloud and Agentforce AI capabilities. |
Every enterprise AI initiative eventually reaches the same crossroads. It isn’t about choosing the best model or investing in more compute — it’s about whether your data is connected well enough for AI to make sense of it.
That challenge sat at the center of “From Fragmented Data to AI-Ready Operations,” an on-demand webinar hosted by LevelShift featuring Charles Betz, VP & Principal Analyst at Forrester; Ananda Paramasivam, CTO of Cornell Lab of Ornithology; and Selva Pandian, Managing Director & Head of Cloud Practice at LevelShift.
The conversation centered on Cornell Lab’s transformation, but the lessons apply to any organization preparing for AI.
Watch the full webinar
A Mission Built on Millions of Small Observations
Cornell Lab of Ornithology exists to help people understand and protect the natural world through birds. Millions of people contribute to that mission through platforms like eBird, Merlin, and Birds of the World — where, as Ananda Paramasivam explained, a single bird sighting “becomes part of the global scientific dataset that can support research, conservation, and migration analysis.”
That’s what makes the Lab unusual: citizen science and world-class research aren’t separate efforts, they continuously feed each other. And behind that mission sits a genuinely enterprise-scale technology footprint — roughly 8 petabytes of data, more than 1,000 years of bird-song recordings preserved in the Macaulay Library, and millions of citizen scientists contributing observations every year.
It’s the kind of infrastructure Ananda Paramasivam didn’t expect to find behind a nonprofit dedicated to birds — cloud platforms, cybersecurity, AI, and CRM modernization, all operating behind the scenes. He joined after three decades at Oracle and a decade leading enterprise technology at Johnson & Johnson, drawn less by the mission alone than by the scale of the problem underneath it: “At some point in your life, you suddenly stop and think, what am I doing? Am I really happy about whatever I’ve been doing?”
The Real Obstacle Wasn’t Volume — It Was Silos
The data itself splits into several very different categories: unstructured media (photos, video, audio), structured citizen-science observations, scientific models derived from that data, and enterprise information covering donors, memberships, finance, and communications.
Betz asked the question most technology leaders eventually face: “Do these all live in fairly separate worlds technically, or does it all come together somewhere?”
It’s a familiar pattern across industries. Systems evolve independently because different teams solve different problems — and the difficulty shows up later, when the business needs those systems to work together.
From Connecting Systems to Connecting People
LevelShift’s engagement with Cornell Lab started narrower: build a Constituent 360 view inside Salesforce Marketing Cloud by unifying data from five business units. It didn’t stay that size for long.
Meaningful personalization meant connecting scientific data, citizen-science activity, memberships, donor records, and commerce data into one ecosystem — a shift that changed how Ananda Paramasivam described the relationship: “They are not data partners. I would call them a transformation partner.”
That broader scope came with a broader focus too: shared capability around identity, governance, cloud architecture, security, privacy, and analytics, rather than simply integrating systems. The goal was never to force everything onto one platform — it was to recognize every constituent as one person with one relationship to the Lab. Someone might identify birds on Merlin, log sightings on eBird, take a Bird Academy course, attend an event, and eventually become a donor. Today those look like separate systems; to Cornell Lab, they’re chapters of the same story.
Recognition That Reflects Where the Industry Is Heading
The webinar also highlighted a broader industry trend. Charles Betz noted that LevelShift now appears in Forrester’s Salesforce Consulting Services Landscape, recognized for its Salesforce Data Cloud and Agentforce AI capabilities alongside leading global consulting firms.
For Selva Pandian, the recognition validates more than fifteen years of Salesforce expertise and continued investment in Data Cloud and Agentic AI.
More importantly, it reflects where enterprise transformation is heading. Organizations aren’t just looking for implementation partners. They’re looking for partners who can build the data foundations AI depends on.
Read more about the Forrester recognition
Why Cornell Didn’t Start With AI
The most counterintuitive moment in the conversation: Cornell Lab wasn’t starting its AI journey — it was sequencing it. Merlin already identifies bird species from photos and audio using machine learning. eBird already turns millions of observations into scientific models used by researchers and conservation groups worldwide.
What was missing wasn’t AI. It was an AI-ready enterprise data foundation on the engagement side of the business. “AI is only useful if the foundation is really a good one — trusted, designed for AI purposes, fully governed and secure. The sequence is very important here,” Ananda Paramasivam said. Selva Pandian made the same point from the delivery side: “AI is only as great as the data behind it.”
That’s why Cornell Lab chose to connect, govern, and secure its donor and member data before expanding AI into that part of the business — even though the organization is already comfortable running AI in production elsewhere.
| Key Takeaways
• Data unification has to come before AI expansion, not after — sequence matters as much as which model you choose.
• Scale isn’t the hard part — fragmentation is. 8 petabytes across formats is manageable; 8 petabytes trapped in disconnected systems is not.
• “Constituent 360” means one person, not one department’s view of that person — every touchpoint (science, learning, giving, membership) is part of one relationship.
• A true transformation partner works on governance, identity, and security alongside system integration, not instead of it.
• Being AI-ready is ultimately a customer-experience question as much as a technical one. |
One Relationship, Not a Collection of Systems
The webinar closed on its most telling moment. Asked what success looks like a few years from now, Ananda Paramasivam didn’t describe a finished implementation or a more advanced AI platform — he described an experience: someone starting on Merlin, moving to eBird, taking a Bird Academy course, attending an event, and eventually becoming a donor, with the Lab recognizing all of it as one relationship rather than five disconnected touchpoints.
That’s the real measure of AI readiness. Not how sophisticated your models are — but whether your data, your systems, and your customer experience tell one connected story.
Frequently Asked Questions
How much data does Cornell Lab of Ornithology manage?
Roughly 8 petabytes, spanning unstructured media (photos, audio, video), structured citizen-science observations, derived scientific data (models, maps, trends), and enterprise data (donor, member, finance, and communications records).
What is a “Constituent 360” view?
A unified view of a person’s full relationship with an organization — in this case, Cornell Lab of Ornithology — that brings together engagement across multiple platforms and business units (e.g., eBird, Merlin, memberships, donations, learning) into one profile, powering Salesforce Marketing Cloud personalization.
Why does data unification come before AI?
Cornell Lab’s CTO argues that AI is only as useful as the data foundation beneath it. The Lab already runs mature AI on its science and mission platforms; the current project focuses on building a trusted, governed data foundation on the engagement side before expanding AI use there.
Ready to make your own data AI-ready? Talk to LevelShift’s team.
