AI agents for loan servicing and collections
Joined June 2018
- Tweets 394
- Following 408
- Followers 302
- Likes 800
160 Photos and videos
šJust another day for proAgent.
24 hours, coast to coast, over $1 million in payments collected.
No hold music, no time zones, no "let me transfer you to someone who can help."
This is what AI agents look like when they're built for collections, not demos.
1
9
Prodigal retweeted
May 19
Shantanu Gangal is teaching AI to read the room on a billion debt-collection calls.
Our cover story on @shantanugangal
yespress.io/shantanu-gangal?ā¦ via Yespress
2
3
74
Prodigal retweeted
What will collections look like in 10 years?
Kyle Cohen says the companies winning then will be the ones investing in AI now. Smart point for anyone thinking about profitability, automation, and long-term strategy.
Watch the clip and full episode: zurl.co/7jC0X
1
2
57
May 13
Prodigal was at Collection and Recovery Solutions (CRS) 2026 last week in Las Vegas.
AI for omnichannel was conversation everyone was having & for good reason. The industry is moving fast & the room felt it.
Good to connect with many people shaping where collections goes next.
1
3
73
Apr 22
Prodigal was at @tcn C3 2026 in Ivins, Utah this week.
Red rock canyons, an outdoor welcome reception, and conversations that made the trip worth it.
Good way to kick off the week!
1
44
Apr 16
For decades, going abroad was the default path for India's best talent.
Sangram Raje, co-founder and CTO of Prodigal, shares his views on why this default is now changing.
@ChannelNewsAsia covered the full story.
channelnewsasia.com/asia/indā¦
1
3
96
Apr 16
The NCBA Executive Experience 2026.
Ponte Vedra, amazing sessions, and even better conversations. Thank you to everyone who stopped by and talked with us this week.
Until next time.
34
Apr 9
Tax season gives collections a short, predictable window where delinquent borrowers are liquid and ready to settle.
Most operations treat it like any other month.
Tax Day is April 15. The window is closing: bit.ly/4t2WbeX
38
Mar 17
Collection agencies don't work every account.
They work the ones that pay best. That's not a flaw, it's how the contingency models are designed.
But the question is what happens to everything else that remains untouched?
Full analysis: bit.ly/476cKO9
1
53
Mar 5
NYCās new rules tightens debt collection.
3 contact attempts per 7 days
Dispute rights: anytime, any channel
60 days to verify debt or it becomes uncollectible
Medical debt requires financial assistance disclosures
What's New York requirement today could be nationwide tomorrow.
1
2
188
Feb 19
Tax refund doesnāt change consumer behavior, refund visibility does.
The moment consumers know a refund is on way, decisions get made.
We break down how refund status affects response rates and why timing matters more than volume: bit.ly/4rUnVkH
1
67
Feb 18
What we shipped:
1ā£proAgent now uses available consumer data to improve how it hears names in real time.
2ā£proAgent evaluates whether spoken name & name on file sound same, for accents, audio distortion & speech variation.
Now, more consumers pass verification on 1st attempt.
1
48
Prodigal retweeted
It started with Clawd
But the name was flawed
Then they did bolt ..
To call it molt š¦
Any case the future is clear ...
So I spoke with @Forbes here
forbesindia.com/article/ai-tā¦
2
3
181
Feb 3
At Prodigal, we build voice AI agents for loan servicing & collections.
A wrong output = compliance issue. So the AI we build on has to be solid.
@Inc42 recently wrote about this, and covered how we use @AnthropicAI.
š Access the full article here inc42.com/features/why-indiaā¦
1
3
129
Jan 28
W-2s arrive next Monday.
A server making $45K gets $8K refund.
An office worker making $80K gets $1.2K.
Why? Tips and overtime deductions (up to $25K each).
Income doesn't predict refunds, what people can claim does.
Stop targeting by income brackets: bit.ly/4rgqfCh
297
Jan 23
Great couple of days at #ARMTech2026 in Dallas.
The user group session brought together a diverse group and sparked some genuinely thoughtful conversations.
And last night, we had dinner with folks across the industry. A good reminder of why in-person connections still matter.
46
Prodigal retweeted
Intelligence layer architecture for collections is what separates scalable AI from risky automation. Shantanu Gangal explains why data context enabling AI decisioning and breaking data silos in receivables are non-negotiable.
Watch here š zurl.co/FbHEW
2
5
82
Jan 20
The IRS starts accepting tax returns next week.
Get your targeting wrong and you'll spend tax season chasing accounts that can't pay.
In part-one of our four-part series we break down what changed with refunds and how to rethink your segmentation: bit.ly/3Nu8MaY
1
2
416
Jan 19
Testing AI agents is hard because every change creates a tradeoff.
Speed up the voice ā interruptions
Fix interruptions ā awkward silence
Reduce latency ā accuracy drops
In consumer finance, assumptions donāt cut it. You have to be absolutely sure: bit.ly/462hGCS
1
1
66
Jan 14
At #ARMTech2026, weāre hosting a User Group Meeting on AI Agents.
Weāll bring Prodigalās on-the-ground experience with lived realities of our customers.
šļøThu, Jan 22 | 2:30-3:30 PM CST
Invite-only.
If youāre attending & like to join, reach out to our team at Booth #42.
1
75