Sanjay Yadav (imatic.ai)

Founder @ the orchestration layer for enterprise Voice AI
Multi-agent voice workflows • CX automation
Bengaluru

[Sanjay Yadav (imatic.ai) @sanjayyadav]

From Operator to Founder: Why I’m Building https://t.co/cagrdAkkXh

I recently took the leap from operating roles to building full-time.

I’m now the Founder of https://t.co/cagrdAkkXh, where we’re building the orchestration layer for enterprise Voice AI.

Voice AI is exploding.
But most companies are still stitching together LLMs, STT, TTS, telephony, workflows, and integrations manually.
It works — but it’s messy.
Hard to scale. Hard to manage. Hard to run in production.

At https://t.co/cagrdAkkXh, we’re building the layer that makes Voice AI production-ready.
Think:
• Multi-agent voice workflows
• Enterprise integrations
• Reliability & monitoring
• Rapid deployment of voice agents

The goal: make deploying Voice AI as simple as deploying software.

Still early days. Building in public.
If you’re working on AI, CX automation, or voice infrastructure — would love to connect.

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@copperheadhq All the best. we applied, didn't get through.

[Sanjay Yadav (imatic.ai) @sanjayyadav]

How come Kolkata owns highest count of NBFC licence in the country ...intriguing.

any thoughts ?

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@banglani Would be good to know, if voice agent acting as an end to end seller or a qualifying layer ? Voice AI + Human in loop would be good number to benchmark ?

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@sridharfyi Would deploy the money for setting up the growth and performance team if D2C or Sales if B2B. Focus is to talk as many customer as possible to find the actual fitment of the product.

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@Kanojiyaaakash1 @SarvamAI @cartesia @cartesia is performing amazingly well ...in most of the Hindi use cases

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@rissa_cao Congrats @rissa_cao . Any plan for native Indian languages??

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@tushaarmehtaa @mayaresearch_ai @SarvamAI @VaaniHQ @bolna_dev As a country we are argumentative. We have thousands of years of oral traditions and 1k+ dialects across the country. When we break the ecosystem between horizontal player and vertical players and hundreds of use cases we prefer to talk ..hence large opportunity in voice ai.

[Sanjay Yadav (imatic.ai) @sanjayyadav]

We are thrilled to launch the video breakdown of https://t.co/xkX8UXfcmz—the premier Agentic Voice AI platform built for the modern enterprise. 🚀

Discover how our infrastructure layer empowers businesses to build monitor, and optimize autonomous voice workflows across business.

[Sanjay Yadav (imatic.ai) @sanjayyadav]

Looks cool Q specific to Instamart

On get_orders response schema
Does it return product name, brand, quantity, price per line item?
Is there a date range filter — can I query month-on-month?
Does it return product images or barcode/EAN?

On search_products
Does the product object include ingredients, nutritional info, or FSSAI data?
Does it return product images (for barcode extraction via vision)?

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@budapp From Orchid to Bud. Infact I built https://t.co/xkX8UXfcmz first verison on Orchid...now looking to explore Bud.

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@ahmeddzeeshan Building https://t.co/xkX8UXfcmz voice agents for enterprise

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@hthieblot https://t.co/xkX8UXfcmz Agentic voice agents for enterprise and SMEs

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@rdominguezibar https://t.co/xkX8UXfcmz agentic voice agents for enterprise and SMEs

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@Techstars https://t.co/xkX8UXfcmz agentic voice agents for enterprise and SMEs.

[Sanjay Yadav (imatic.ai) @sanjayyadav]

https://t.co/wdG5c7I3yZ

[Sanjay Yadav (imatic.ai) @sanjayyadav]

@hthieblot Building https://t.co/xkX8UXfcmz,voice agents and orchestration platform for enterprise and SMEs

[Sanjay Yadav (imatic.ai) @sanjayyadav]

The voice stack is becoming what the data center was to cloud.

Unsexy. Critical. Defensible.

The winners won't have the smartest model.

They'll be the ones who treated real-time orchestration as a first-class engineering problem.

If this was useful, repost or share your thoughts.

[Sanjay Yadav (imatic.ai) @sanjayyadav]

Most Voice AI companies are optimizing the wrong thing.

They're swapping models. Fine-tuning. Chasing 2% accuracy gains.

The bottleneck isn't intelligence.

It's latency. 🧵

[Sanjay Yadav (imatic.ai) @sanjayyadav]

Infrastructure compounds differently than AI:

Models get copied. Infrastructure doesn't. Models deprecate.

Latency gains stack. Models are a cost line. Infrastructure is a moat.