Draft, written by the Bee

Neon

Pre-meeting dossier ·

Written by the Bee, read it before you send it

KPIMore first meetings turn into real deals, on less preparation time

selling on-prem storage

Why now

Since closing the ~$1B Databricks acquisition in May 2025 and rebranding as Lakebase Postgres inside Databricks, Neon is now integrating into a much larger platform org where internal tooling sprawl and cross-team coordination typically spike.

Lead with

Unified AI workspace ("Enterprise Brain")

Best fit

Private-cloud deployment options with GDPR-compliant, non-training data handling medium

01 · The company

Neon today

Industry
Cloud database infrastructure (serverless PostgreSQL)
Size
~50-150 employees (per 2024 funding coverage); now operating as a Databricks business unit
HQ
Menlo Park, California
Footprint
Runs on AWS regions globally · San Francisco-based parent (Databricks) · Serves developers and enterprises worldwide via cloud

Neon builds serverless Postgres that separates storage from compute, letting databases autoscale, branch like Git, and pause to zero cost when idle. Founded in 2021, it was acquired by Databricks in 2025 and now operates as "Lakebase Postgres" inside the Databricks platform, positioned as the backend for both human developers and AI agents.

Recent moves
  • Neon.com now markets itself as "the backend for apps and agents," bundling Postgres, Auth, Functions, Storage, and an AI Gateway (site copy, 2026)
  • Databricks closed its ~$1B acquisition of Neon in May 2025
  • Neon reported over 80% of new databases on its platform are now created by AI agents rather than humans, up from 30% a year earlier
  • Neon scaled to 700,000+ databases created, up from ~6,000 previously, per 2026 comparison coverage
Public pain
  • Sacra estimated Neon's ARR at only ~$25M at time of a ~$1B acquisition price, implying pressure to prove enterprise monetization
  • No reserved-capacity pricing exists, so a 10x traffic spike produces a 10x compute bill for customers
  • Region lock-in: Neon databases are limited to a small number of AWS regions
Strategic bets
  • Committed to a
  • auth, functions, storage, AI gateway
  • platform strategy — "Lakebase Postgres, Auth, Functions, Storage, and an AI Gateway" (site positioning)
4 sources
Neon — Postgres backends for apps and agentsDatabricks Agrees to Acquire NeonNeon revenue, funding & news | SacraNeon vs Supabase 2026 - Tech Insider
02 · Lead with this

Positioning

NSOffice.AI gives enterprise teams a single prompt interface to search, automate, and execute across their scattered business systems instead of stitching together a dozen point tools. For a company scaling fast, that means employees stop losing hours hopping between systems and start executing straight from one governed workspace.

03 · What to pitch

Pitches

Unified AI workspace ("Enterprise Brain")

Give Neon's Finance, Sales, and Ops staff one prompt interface instead of juggling newly-merged Databricks systems on top of their own. It cuts the ramp-up cost of post-acquisition tool sprawl.

Multi-model access layer routing queries across 10+ AI models

As Neon positions itself as backend for AI agents, internal teams get the same multi-model flexibility for their own workflows without picking one vendor lock-in. This keeps pace with the AI-agent-first narrative Neon is already selling externally.

04 · How to run the room

Conversation arc

open
Establish relevance post-acquisition

“Congrats on the Databricks close — integrating into a platform that size usually means your internal teams are juggling more systems, not fewer.”

discover
Surface internal tooling pain from the integration

“As Neon folds into Databricks' processes, are your Finance, Sales, or Ops teams having to bounce between more disconnected tools than before?”

present
Introduce the single-prompt workspace as the fix

“NSOffice.AI puts one prompt interface in front of your teams' systems instead of another app to learn on top of Databricks' stack.”

objection
Address stack-fit skepticism

“I hear you — this isn't about replacing Postgres or your data layer, it's about the layer your people work in above it.”

close
Get a scoped next step

“Can we set up a working session with whoever owns internal tooling decisions post-merger to map where the friction actually is?”

05 · Discovery

Ask these

  1. Q1Now that Neon is a Databricks business unit, which teams are dealing with duplicated or newly-added systems since the merger?
  2. Q2Your docs mention PgBouncer handles up to 10,000 client connections — is engineering support load or internal ops tracking scaling with that kind of growth?
  3. Q3With 700,000+ databases now created on the platform, has that growth created new internal reporting or cross-functional coordination overhead?
  4. Q4Region lock-in to 8 AWS regions was flagged as a constraint — does that limitation create manual workarounds your team has to track internally?
06 · Where you fit

Feasibility

Private-cloud deployment options with GDPR-compliant, non-training data handling
product
medium
Neon's own pain point of AWS region lock-in and PrivateLink-based private networking suggests some latent interest in deployment flexibility and data control, though this is a stretch since NSOffice's offering is a productivity layer, not storage infrastructure.
Governance layer with logging, audit trails, RBAC, SSO, and data residency controls for compliance teams
product
low
Neon operates as an infrastructure/database vendor, not a knowledge-worker productivity buyer, so governance-layer positioning has weak fit given the meeting context is on-prem storage, not employee workflow compliance.
Pre-built integrations with 15+ business systems (email, drives, CRM, ERP, HRMS, shared folders)
product
low
Neon's tech estate is AWS/Postgres/Databricks-centric with no CRM, ERP, or HRMS signals surfaced in research, so this integration set doesn't map to their stack.
Unified AI workspace ("Enterprise Brain") for search, task automation, content generation, and data analysis via a single prompt
solution
low
There is no signal that Neon's internal teams face the fragmented-tool problem this solves; the research is about their product/infrastructure, not internal ops tooling.
Autonomous workflow execution ("Advise to Execute") for functions like Finance, Sales, Marketing, and Ops
solution
low
No named signal ties Neon's Finance/Sales/Ops functions to workflow automation needs; this is a generic fit without a specific anchor.
07 · Technology estate

What they appear to run

Inferred stack
PostgreSQL (core engine)AWS (hosting regions)PgBouncer (connection pooling)Databricks Lakebase (parent platform integration)S3-compatible object storagePrivateLink (private networking)
Key vendors
AWSDatabricksVercelGitHub
Signals
  • Neon's own docs describe built-in PgBouncer-based connection pooling handling up to 10,000 client connections
  • Site lists integrations/partnerships with Vercel, Netlify, GitHub Actions, and Cursor plugins
  • Homepage confirms Neon databases only run on 8 AWS regions and offer PrivateLink for private networking
3 sources
Neon revenue, funding & news | SacraNeon — Postgres backends for apps and agentsNeon LinkedIn