Shriram Finance
less than 3–4 hours — Onboarding Turn-Around-Time (TAT)
Independently reportedBanking & finance · Process automation · Classical ML · In production
Process automation
Basata · Hamming
Vendor-reportedHealthcareUnited StatesScaledLLM
Vendor-reported. The customer is named and the numbers are quoted from the source page, but the account comes from the vendor. No independent confirmation.
Reported by hamming.ai (vendor self-report). Checked against the source page on 2026-09-05. 2 of 3 figures below appear on that page word for word; the rest are marked.
The problem
Manual testing of voice AI agents was slow, unscalable, and couldn't cover diverse scenarios, leading to high-stakes risks in healthcare.
What was deployed
- Vendor
- Hamming
- Products
- Hamming
- Technique
- LLM
- Build or buy
- Bought off the shelf
- Deployment
- API
- Data used
- agent configuration, evaluator descriptions and failure analysis, persona library with regional voices and speech patterns
- Scale
- scales with every new customer
What changed
Each row is quoted from the source. Figures we could not find on the page in those words are marked — they are kept, not deleted, so you can judge them.
manual testing time
“Reduction in manual testing time 83%”
Not found on the source page in these words
15 minutestest execution timefrom 90 minutesper cycle
“Testing time dropped from 90 minutes of manual phone calls to 15 minutes of automated concurrent execution.”
Quoted word for word from the source
6x fasterTesting cycle speed
“6x Faster Testing Cycles”
Quoted word for word from the source
Entire test suite runs simultaneouslytest parallelismfrom One person, one call at a time
“With Hamming, the entire test suite runs concurrently.”
Quoted word for word from the source
Programmatically converts it into test casestest case generationfrom Manually writing and maintaining test cases
“Automated test case generation via API”
Quoted word for word from the source
Automated agentic feedback loop from evaluator resultsprompt optimizationfrom Manual prompt engineering: trial and error
“Agentic Feedback Loops for Prompt Optimization”
Quoted word for word from the source
“I find talking to the agents completely socially exhausting. The platform offers all of these different personas that the team is not able to replicate. We want to make sure that our diverse customer base is represented in the agents themselves.”
Difficulties and limits
- 1.5 hours of manual phone calls per test cycle, growing with every new customer
- One person, one call at a time, no parallelism possible
- Cannot simulate diverse patient accents, demographics, or speech patterns
- Socially exhausting, cognitive energy spent on calling instead of building
- High-stakes healthcare context where failures mean patients don't get care
Sources
- hamming.ai — vendor self-report, 2026-03-23
Similar deployments
GE Aerospace
Independently reportedManufacturing · Process automation · Classical ML · Scaled
Apollo Tyres
under 10 minutes — Time for root cause analysis
Independently reportedManufacturing · Process automation · AI agents · In production
Broadcom · Moveworks
88% — autonomous IT resolution rate
Independently reportedTechnology & software · Process automation · NLP · Scaled
icare · SAS
increased by 15% — accuracy
Independently reportedInsurance · Process automation · Classical ML · Scaled · 2019
Amazon · Amazon
10% more — shopper spend
Independently reportedRetail & e-commerce · Process automation · Computer vision · Scaled · 2024