Pinned Tweet
Voice, innovation, and ideas are about to align in perfect harmony under one roof!
Take your seats for the biggest orchestration ever... this November, in Bengaluru.
Symphony is coming alive! Stay tuned!
Take your seats for the biggest orchestration ever... this November, in Bengaluru.
Symphony is coming alive! Stay tuned!
1
6
20
2,854
Happy Diwali from Team @bolna_dev! 🪔
Last Diwali, five of us celebrated around a single table. This year, we needed a considerably larger room for a full house (the group photo now needs a wide-angle lens).
Slack went quiet for an evening, and nobody missed it.
Wishing your happy as ours was! ✨
Last Diwali, five of us celebrated around a single table. This year, we needed a considerably larger room for a full house (the group photo now needs a wide-angle lens).
Slack went quiet for an evening, and nobody missed it.
Wishing your happy as ours was! ✨
0
2
16
664
bolna retweeted
Excited to launch DECIDE: The first ever benchmark for audio decision models made from production calls data.
Submit your model's API endpoints and evaluate your model on the benchmark.
Thanks to the folks at @bolna_dev for the collaboration.
Submit your model's API endpoints and evaluate your model on the benchmark.
Thanks to the folks at @bolna_dev for the collaboration.
Introducing DECIDE: a benchmark for audio decisions in voice agents, in association with @bolna_dev
A voice agent doesn’t just transcribe. It decides: speak or wait, route the call, pick a tool, switch language. Text can’t tell you if the caller sounded annoyed, interrupted, or DECIDE tests exactly that, on real production calls.
→ 4,867 decision questions from 1,081 real agent-caller clips (3 to 25 seconds)
→ 32 tasks across 6 families: turn-taking, graph routing, tool decisions, language switching, caller emotion, business intent
Direct-audio models and STT + text models (20 in total) are scored on the same questions, so you can see what audio actually adds, and what it costs in dollars and milliseconds.
Submit your model endpoint and get scored; no invite needed.
Website: benchmark.tazolabs.com
Samples: huggingface.co/datasets/tazol…
1
1
9
403
Voice Agents do more than understand what you say. They decide what to do next.
Introducing DECIDE, a benchmark for evaluating those decisions on real production calls.
benchmark.tazolabs.com
Introducing DECIDE, a benchmark for evaluating those decisions on real production calls.
benchmark.tazolabs.com
Introducing DECIDE: a benchmark for audio decisions in voice agents, in association with @bolna_dev
A voice agent doesn’t just transcribe. It decides: speak or wait, route the call, pick a tool, switch language. Text can’t tell you if the caller sounded annoyed, interrupted, or DECIDE tests exactly that, on real production calls.
→ 4,867 decision questions from 1,081 real agent-caller clips (3 to 25 seconds)
→ 32 tasks across 6 families: turn-taking, graph routing, tool decisions, language switching, caller emotion, business intent
Direct-audio models and STT + text models (20 in total) are scored on the same questions, so you can see what audio actually adds, and what it costs in dollars and milliseconds.
Submit your model endpoint and get scored; no invite needed.
Website: benchmark.tazolabs.com
Samples: huggingface.co/datasets/tazol…
0
6
20
492
bolna retweeted
Introducing DECIDE: a benchmark for audio decisions in voice agents, in association with @bolna_dev
A voice agent doesn’t just transcribe. It decides: speak or wait, route the call, pick a tool, switch language. Text can’t tell you if the caller sounded annoyed, interrupted, or DECIDE tests exactly that, on real production calls.
→ 4,867 decision questions from 1,081 real agent-caller clips (3 to 25 seconds)
→ 32 tasks across 6 families: turn-taking, graph routing, tool decisions, language switching, caller emotion, business intent
Direct-audio models and STT + text models (20 in total) are scored on the same questions, so you can see what audio actually adds, and what it costs in dollars and milliseconds.
Submit your model endpoint and get scored; no invite needed.
Website: benchmark.tazolabs.com
Samples: huggingface.co/datasets/tazol…
A voice agent doesn’t just transcribe. It decides: speak or wait, route the call, pick a tool, switch language. Text can’t tell you if the caller sounded annoyed, interrupted, or DECIDE tests exactly that, on real production calls.
→ 4,867 decision questions from 1,081 real agent-caller clips (3 to 25 seconds)
→ 32 tasks across 6 families: turn-taking, graph routing, tool decisions, language switching, caller emotion, business intent
Direct-audio models and STT + text models (20 in total) are scored on the same questions, so you can see what audio actually adds, and what it costs in dollars and milliseconds.
Submit your model endpoint and get scored; no invite needed.
Website: benchmark.tazolabs.com
Samples: huggingface.co/datasets/tazol…
89
203
880
1,566,630
Bengaluru, it’s time to plan your day! 🚀
The agenda for the big day, Symphony: India’s flagship Voice AI Summit, is finally out!
Across Sur and Abhyas, the day runs on two stages at once. Sur, our main stage, brings together the people shaping voice AI in India: global AI speech models for India’s languages, and the investors betting on what comes next. Abhyas, the build stage, is for those who want to see how it all comes together, with product showcases, sessions and the Buildathon finals.
Team @bolna_dev will also be unveiling what our team has been quietly building for months, so do stay for that.
Grab a coffee, pick your sessions, and maybe pack your comfy shoes for running between stages. Link to the full agenda: bolna.ai/symphony/agenda
P.S. You’ll spot a special performance on the agenda, and it’s the part Team Bolna is most excited about. Keeping it under wraps has been killing us, so stay tuned.
The agenda for the big day, Symphony: India’s flagship Voice AI Summit, is finally out!
Across Sur and Abhyas, the day runs on two stages at once. Sur, our main stage, brings together the people shaping voice AI in India: global AI speech models for India’s languages, and the investors betting on what comes next. Abhyas, the build stage, is for those who want to see how it all comes together, with product showcases, sessions and the Buildathon finals.
Team @bolna_dev will also be unveiling what our team has been quietly building for months, so do stay for that.
Grab a coffee, pick your sessions, and maybe pack your comfy shoes for running between stages. Link to the full agenda: bolna.ai/symphony/agenda
P.S. You’ll spot a special performance on the agenda, and it’s the part Team Bolna is most excited about. Keeping it under wraps has been killing us, so stay tuned.
1
2
9
315
Enterprises that answer calls across Ladakh to Kanyakumari or Gujarat to Manipur shouldn't be limited to agents who only know how to answer in one language. Whether it's a bank, a healthcare facility or a national university, the caller on the line deserves an agent that meets e solve? A dedicated prompt for every language.
Except that always becomes too tedious to build with each add-on.
Adapting to voice agents was supposed to make this easier. Instead, adding new languages only meant adding another version of the agent with an additional prompt to write and test, with each linguistic version drifting apart and getting lost in translation.
@bolna_dev's latest update on Canvas takes away that friction for enterprises with just one edit.
Draft your agent’s logic once, and it holds across every language the agent is supposed to speak. Only the lines the agent says out loud change from one language to the next. Bolna translates them automatically, and teams can still fine-tune the wording so it sounds natural to local callers.
Adding a new language takes just a click, and not a rewrite. What you end up with is multiple agents fluent in the language your customers speak without an added effort of a dedicated prompt for each one of them.
Go try your new Canvas on Bolna now!
Except that always becomes too tedious to build with each add-on.
Adapting to voice agents was supposed to make this easier. Instead, adding new languages only meant adding another version of the agent with an additional prompt to write and test, with each linguistic version drifting apart and getting lost in translation.
@bolna_dev's latest update on Canvas takes away that friction for enterprises with just one edit.
Draft your agent’s logic once, and it holds across every language the agent is supposed to speak. Only the lines the agent says out loud change from one language to the next. Bolna translates them automatically, and teams can still fine-tune the wording so it sounds natural to local callers.
Adding a new language takes just a click, and not a rewrite. What you end up with is multiple agents fluent in the language your customers speak without an added effort of a dedicated prompt for each one of them.
Go try your new Canvas on Bolna now!
0
0
5
219
Welcome to the ensemble, Sarvam and Vobiz AI.
We’re so excited to have you at Symphony: India’s Flagship Voice AI Summit.
India’s voice AI builders, founders and operators will all be under one roof, so make sure you’re in it too.
Link to get your golden ticket: lnkd.in/p/gf57KzMR
🗓️ November 20, 2026
📍 The Lalit Ashok, Bengaluru
We’re so excited to have you at Symphony: India’s Flagship Voice AI Summit.
India’s voice AI builders, founders and operators will all be under one roof, so make sure you’re in it too.
Link to get your golden ticket: lnkd.in/p/gf57KzMR
🗓️ November 20, 2026
📍 The Lalit Ashok, Bengaluru
1
1
6
160
bolna retweeted
professional yappers of bangalore, your time has come 😭🎙️
we're doing Build Out Loud with @WisprFlow × @bolna_dev , a hackathon where you build voice AI agents without typing a single prompt. just talk, build, break things, fix them, repeat.
📍 blr | nov 15 | 12–5 pm
👥 solo or teams of up to 4
your yap might actually win you something 👀
luma.com/1a2rjj9r
we're doing Build Out Loud with @WisprFlow × @bolna_dev , a hackathon where you build voice AI agents without typing a single prompt. just talk, build, break things, fix them, repeat.
📍 blr | nov 15 | 12–5 pm
👥 solo or teams of up to 4
your yap might actually win you something 👀
luma.com/1a2rjj9r
4
2
45
1,318
Can you build something without typing a single line?
@WisprFlow and @bolna_dev are teaming up to get builders together for Build Out Loud, a voice only Buildathon, putting this question to test and a lil’ something to get you hyped for Symphony.
Every build you’ve shipped this one starts with your voice.
If you have got a blank page staring back at you, we’ve got you, try an AI recruiter, a multilingual support agent, an outbound agent that qualifies leads or a feedback agent that runs surveys.
Link to register: luma.com/1a2rjj9r
🗓️ 15th November | 12 PM to 5 PM
👥 Solo or teams of up to 4
PS: Loud ideas are welcomed, wispr-ed ones even more so!
@WisprFlow and @bolna_dev are teaming up to get builders together for Build Out Loud, a voice only Buildathon, putting this question to test and a lil’ something to get you hyped for Symphony.
Every build you’ve shipped this one starts with your voice.
If you have got a blank page staring back at you, we’ve got you, try an AI recruiter, a multilingual support agent, an outbound agent that qualifies leads or a feedback agent that runs surveys.
Link to register: luma.com/1a2rjj9r
🗓️ 15th November | 12 PM to 5 PM
👥 Solo or teams of up to 4
PS: Loud ideas are welcomed, wispr-ed ones even more so!
1
4
12
968
2 weeks to showtime!
Today we welcome @BizTruecaller , @Exotel and @KalpaLabsAI to the ensemble, joining @GoogleCloud_IN . All four, Diamond Sponsors by tier (precious by nature).
Last week, Mumbai got a glimpse of what Symphony has to offer. Bengaluru, we hope you’re ready ce AI in India has witnessed.
Link to reserve a spot and get the tickets- bolna.ai/symphony
📅 November 20, 2026
📍 The Lalit Ashok, Bengaluru
Today we welcome @BizTruecaller , @Exotel and @KalpaLabsAI to the ensemble, joining @GoogleCloud_IN . All four, Diamond Sponsors by tier (precious by nature).
Last week, Mumbai got a glimpse of what Symphony has to offer. Bengaluru, we hope you’re ready ce AI in India has witnessed.
Link to reserve a spot and get the tickets- bolna.ai/symphony
📅 November 20, 2026
📍 The Lalit Ashok, Bengaluru
0
1
7
264
We’re kicking off Diwali month with something we’ve been excited to share: Bolna Partners Program 🪔
It’s built for agencies, consultants and product teams who want to bring voice AI to their customers.
Here’s how it works, partners deliver Voice AI under their own brand, and neath it. You own the customer relationship, the implementation and the support and we handle the infrastructure.
What partners get:
• An API-first platform built for white labeling
• Infrastructure that already handles 1M+ calls a day
• Preferred rates from day one, with no minimum volume to unlock them
We know this model works because it’s already working. 4 of our top 10 customers came to us through partners. Now we’re putting a proper program behind it.
If you’ve been looking for the right Voice AI partner, this is your sign.
Apply here : bolna.ai/partners
It’s built for agencies, consultants and product teams who want to bring voice AI to their customers.
Here’s how it works, partners deliver Voice AI under their own brand, and neath it. You own the customer relationship, the implementation and the support and we handle the infrastructure.
What partners get:
• An API-first platform built for white labeling
• Infrastructure that already handles 1M+ calls a day
• Preferred rates from day one, with no minimum volume to unlock them
We know this model works because it’s already working. 4 of our top 10 customers came to us through partners. Now we’re putting a proper program behind it.
If you’ve been looking for the right Voice AI partner, this is your sign.
Apply here : bolna.ai/partners
1
2
8
317
join us: luma.com/6uhcmz3l
Build, buy, or wait? What's the right call for voice AI? 🎙️
On Nov 4, we're getting Devansh Ghatak (Co-founder & CTO, @SimplismartHQ) and @xan_ps (Co-founder & CTO, @bolna_dev) together to talk about this.
What's worth building in-house, what makes more sense to buy, and what has chanAI in production?
Bring your questions too!
📅 Nov 4 | 5–6 PM IST
(registration link shared in comments)
0
0
6
308
bolna retweeted
Build it. Buy it. Or wait for it?
🗓 Wed, 4 Nov · 5–6 PM IST
Register: luma.com/6uhcmz3l
Every CTO is facing the voice AI question this year. Get it wrong and you lose a year of engineering, get locked into a stack, and watch customers hang up on a slow agent.
So we're n voice AI at scale on the same stage:
Devansh Ghatak, Co-founder & CTO, @SimplismartHQ
Prateek Sachan, CTO, @bolna_dev (YC F25)
What we'll get into:
→ Which build-vs-buy assumptions broke in the last 12 months
→ What two of India's biggest voice platforms actually see in production
→ Who owns voice, what gets measured, and how pilots become scale
→ The 2027 call: does the voice stack collapse into one model?
→ Your questions, live!
🗓 Wed, 4 Nov · 5–6 PM IST
Register: luma.com/6uhcmz3l
Every CTO is facing the voice AI question this year. Get it wrong and you lose a year of engineering, get locked into a stack, and watch customers hang up on a slow agent.
So we're n voice AI at scale on the same stage:
Devansh Ghatak, Co-founder & CTO, @SimplismartHQ
Prateek Sachan, CTO, @bolna_dev (YC F25)
What we'll get into:
→ Which build-vs-buy assumptions broke in the last 12 months
→ What two of India's biggest voice platforms actually see in production
→ Who owns voice, what gets measured, and how pilots become scale
→ The 2027 call: does the voice stack collapse into one model?
→ Your questions, live!
1
5
8
416
Build, buy, or wait? What's the right call for voice AI? 🎙️
On Nov 4, we're getting Devansh Ghatak (Co-founder & CTO, @SimplismartHQ) and @xan_ps (Co-founder & CTO, @bolna_dev) together to talk about this.
What's worth building in-house, what makes more sense to buy, and what has chanAI in production?
Bring your questions too!
📅 Nov 4 | 5–6 PM IST
(registration link shared in comments)
On Nov 4, we're getting Devansh Ghatak (Co-founder & CTO, @SimplismartHQ) and @xan_ps (Co-founder & CTO, @bolna_dev) together to talk about this.
What's worth building in-house, what makes more sense to buy, and what has chanAI in production?
Bring your questions too!
📅 Nov 4 | 5–6 PM IST
(registration link shared in comments)
2
4
13
1,049
If you’re building an AI voice agent, tweaking it shouldn’t be a two-hour long tedious process that has you spending hours working out what needs to change.
But here’s the good part, you can now hand that tedious part over to AI Edit on @bolna_dev.
Open your Canvas, just point ght, and let it do its job.
You'd have your voice agent ready exactly the way you imagined.
Try it on while it’s free in beta, link in threads!
But here’s the good part, you can now hand that tedious part over to AI Edit on @bolna_dev.
Open your Canvas, just point ght, and let it do its job.
You'd have your voice agent ready exactly the way you imagined.
Try it on while it’s free in beta, link in threads!
14
2
27
561
The next set of voices are here.
We’re excited to welcome @DhawanAnuja from @Exotel, @pshishodiaa from @KalpaLabsAI, and @madhkrish from @VahanInc to the panel.
The stage is coming together, and so are the conversations we’ve been waiting for.
Link to get the tickets is in the em live at Symphony on November 20th at The Lalit Ashok!
We’re excited to welcome @DhawanAnuja from @Exotel, @pshishodiaa from @KalpaLabsAI, and @madhkrish from @VahanInc to the panel.
The stage is coming together, and so are the conversations we’ve been waiting for.
Link to get the tickets is in the em live at Symphony on November 20th at The Lalit Ashok!
0
1
11
371
We said this could be your best shot at a direct interview with @ycombinator .
Then 2,000+ of you pulled up, so the Buildathon had to level up!
Thanks to your response and our Buildathon Partner @plivo , @bolna_dev 's Buildathon has gotten bigger:
• ₹5L in cash prizes (more rted with)
• Top 5 teams pitch live on the Abhyas stage at Symphony, and we’ll fly you to Bengaluru.
• Top 50 teams get complimentary Symphony tickets
• The winner gets an interview with Y Combinator (still the crown jewel)
The clock is ticking faster than ever and there’s only 72 hours left to shoot your shot.
Link to register: bolna.ai/symphony/build…
Then 2,000+ of you pulled up, so the Buildathon had to level up!
Thanks to your response and our Buildathon Partner @plivo , @bolna_dev 's Buildathon has gotten bigger:
• ₹5L in cash prizes (more rted with)
• Top 5 teams pitch live on the Abhyas stage at Symphony, and we’ll fly you to Bengaluru.
• Top 50 teams get complimentary Symphony tickets
• The winner gets an interview with Y Combinator (still the crown jewel)
The clock is ticking faster than ever and there’s only 72 hours left to shoot your shot.
Link to register: bolna.ai/symphony/build…
1
4
35
1,919