Case study Member onboarding · Social investment network

4x-ing AVPN's member onboarding throughput, in 5 weeks.

How Asia's largest network of social investors rebuilt member onboarding into a single system, and quadrupled throughput without adding headcount.

Results AVPN × realfast
Monthly throughput
~4×
Same team: 71 onboarded in month one vs the usual 16 to 20
Tools condensed
6 → 1
Six tools condensed into one system, submission to active membership
Due diligence
~50 sec
AI-led, per application; approval activates the member

Live in production on 1 July, five weeks after kickoff.

Six informal, manual handoffs became one structured flow. An application used to be copy-pasted across Gravity Forms, email, Airtable, Salesforce, payments and spreadsheets; it now runs as a single orchestrated handoff in one system.

realfast redefines processes rather than just automating them. Their deep understanding of both AI and business enabled us to consolidate six systems into one and cut approval times to ~50 seconds through AI-led due diligence. As a result, we’re handling four times the application volume without adding headcount. Working with them felt less like hiring a vendor and more like gaining true team members.

Naina Subberwal Batra Naina Subberwal Batra CEO, AVPN
The problem

Why they called

AVPN is Asia's largest network of social investors, a base that only grows as more organizations join. But every new member had to be carried by hand across six tools. Applications came in through Gravity Forms as email, were copied into Airtable by hand, pushed through Salesforce, run through a payments tool and tracked in spreadsheets, and re-keyed at each step. Someone triggered the due-diligence checks manually, one prompt at a time, and approvals bounced back and forth over email.

On that setup the market team could approve 16 to 20 members in a typical month, and handling more meant hiring more. For a network whose value compounds with its size, onboarding had become the ceiling on growth.

Why it mattered

The stakes

The free membership tier was set to grow, and a larger network is a more valuable one. But a free application is not a lighter one. It goes through the same verification, the same due diligence and the same approval as a paid member. Growing the free tier grows the number of applications, and each one costs the team just as much to process. Onboarding capacity, not membership fees, was the real constraint on growth.

Our approach

Redefined the process, not just automated it

We did not automate the six-tool relay, we replaced it. One system now carries an application from submission to active membership, and the decision stays with the team.

One system, end to end.

Intake, records, payments and membership state live in one place, so the six informal handoffs become a single structured flow, with nothing re-keyed and nothing lost between tools.

AI-led due diligence.

Checks that used to be triggered by hand, one prompt at a time, now run on their own in about 50 seconds, and approval activates the member with no further steps.

Configurable by the team.

The Vantage and Discovery intake forms, platform copy, email templates and staff role assignments are all editable by AVPN directly, with no development request.

The decision stays with AVPN.

The platform assembles and checks the application; a person approves.

Member onboarding

One system, submission to active membership

Nothing re-keyed between tools, and nothing waiting on someone to remember to trigger it.

Apply Vantage or Discovery intake form
Verify Records and payment in one place
Due diligence AI-led, about 50 seconds
Approve A person decides, and the member goes active
Configurable by AVPN Both intake forms, platform copy, email templates and staff role assignments — all editable by the team directly, with no development request.
Only one stage is new work for a machine. Apply, verify and approve were always steps; due diligence was the one that had to be triggered by hand, one prompt at a time, and it is the one that moved.
MEMBER ONBOARDING Six tools, one system BEFORE · SIX TOOLS, SIX HANDOFFS Gravity Forms Email Airtable Salesforce Payments tool Spreadsheets Re-keyed at each step. Due diligence triggered by hand, one prompt at a time. Approvals bounced over email. 16–20 a month Handling more meant hiring more. REBUILT AFTER · ONE SYSTEM, ONE FLOW APPLY Vantage or Discovery intake form VERIFY Records and payment in one place DUE DILIGENCE AI-led, about 50 seconds APPROVE A person decides, and the member goes active One structured flow. Nothing re-keyed, nothing lost between tools. The approval still belongs to AVPN. 71 in month one Same team. No headcount added.
the consolidation as drawn: realfast · AVPN member onboarding
The build

Kickoff to production in five weeks

Five weeks, kickoff to production.

Live on 1 July, with the safety and audit trail built in from the start.

A full month in production.

The first complete month of onboarding ran end to end on the platform: 71 members onboarded, with discovery and due diligence running on their own.

Kickoff to production

Five weeks, then a month of proof

A working prototype every week, and dependencies called out a week ahead, so AVPN could plan their side in time.

Kickoff Scope agreed across the three teams onboarding touched
Weeks 1–5 Build, reviewed in place each week rather than agreed from a spec
1 July 2026 Live in production, with the safety and audit trail built in from the start
July The first complete month: 71 members onboarded, diligence running on its own
The month after go-live is the part that counts. Five weeks is how fast it was built; July is the evidence it holds — a full month of onboarding run end to end on the platform.
71 members in month one

against the 16 to 20 a typical month allowed on the six-tool relay — and on the same team.

Before — a typical month
16–20
After — the first full month
71

Members onboarded per month. Before, the 16 to 20 the market team could approve across the six-tool relay; after, 71 in the first complete month with the platform running end to end — July 2026, with no headcount added.

How we worked with the team

Why it landed at this pace

Onboarding cut across three teams and business functions, so how we worked mattered as much as what we built.

Kept 10+ stakeholders moving together.

Onboarding touched more than ten people across three teams; we ran it as one shared effort and kept everyone comfortable at this pace.

The prototype was the conversation.

A high-fidelity prototype let people experience the real flows and comment before anything shipped, which pulled in feedback from stakeholders we never spoke to directly, including their branding team. Around 30 comments came back this way.

Interactive docs for every customization.

Rather than triggering each flow, the team reviewed and changed things in place — the email templates, who gets which role, and the data shown on the member page — all on their own time.

A UAT tracker that ran itself.

One interactive tracker held the portal links, test credentials, guidelines and self-tracked coverage, so testing and sign-off were easy.

Built to hand over.

Reusable components they reconfigure themselves, a maintenance guide (API keys, copy, templates, member roles) and deployment and verification pipelines, so they change production confidently without breaking anything, and without us.

Dependencies flagged a week ahead.

We shipped a working prototype every week and called out dependencies and blockers early, so the team could plan their side in time.

In five weeks, realfast helped my team dramatically increase how many members we could onboard — with the same headcount. That’s the outcome; how they got there is the real differentiator. Everyone says they’re “AI-native” — realfast actually is, and you feel it in the way they work. Instead of long specs and static documents, most of what we reviewed together was a working prototype we could react to live. That’s how they understood our onboarding and re-engineered it with AI faster than a traditional build would have allowed. And it’s live, doing that today.

Sharman Pandian Sharman Pandian Lead, Systems and Operations, AVPN
Partnership

What AVPN owned

AVPN owned what only it could: who qualifies as a member, the tiers, and the approval itself. realfast built the platform that carries the application and runs the diligence, but the decision, and the ability to reshape the process, stayed with the team. The upshot is simple: handling more members no longer requires more people.

Talk to us

Does growth still mean hiring?

If growth in your organization still means hiring more people to push work across more tools, that is the pattern we rebuild. Talk to us.

Let's go

Thanks, we're thrilled you're here. We'll be in touch shortly to set up your session.

Something went wrong

We couldn't submit your request. Please try again, or email us at hello@realfast.ai.

Figures as delivered and verified in production — the first full month of member onboarding run end to end on the platform, July 2026.

Take this case study with you

Save a clean, print-ready PDF to share internally or send to your team.