The Role
Triage daily technical support tickets for voice-AI agents, perform root-cause analysis using transcripts, logs, and SQL, debug APIs/webhooks and prompts, fix issues when possible, escalate with evidence and reproducible steps, and document failure modes for the knowledge base.
Summary Generated by Built In
About Bolna
About the role
What you'll own
What you bring
Nice to have
Success in this role looks like
Why Bolna
Bolna is Voice AI infrastructure built for India - and now for the world. We help businesses deploy intelligent voice agents that can call, converse, and convert in any language, at scale. From collections to customer support to sales, our agents handle millions of conversations so humans don’t have to.
We’re a YC F25 company, backed by General Catalyst, with 1,050+ paying customers and growing fast. Our team of ~40 is based in Bengaluru.
Bolna's agents run across a stack of ASR, LLM, TTS, and telephony providers, at volume, across dozens of clients - which means something is breaking somewhere almost every day. A call that failed. A prompt that didn't do what a client expected. A webhook that stopped firing. This role is the first line on that queue: for every ticket, you have to work out what actually broke, why, and whether you can fix it yourself.
You won't usually be able to look up the answer in a known-issues doc, because a lot of what breaks at Bolna hasn't broken in quite that way before. You're expected to reason from first principles: pull the transcript, pull the logs, form a hypothesis, test it, fix what's yours to fix, and write a sharp handoff for what isn't. Nobody's handing you a script.
If you like puzzles more than tickets, this is a good fit.
What you'll own
- Ticket triage and root-cause analysis. Work the daily queue of technical support issues. Diagnose the actual root cause using transcripts, call logs, and system metadata before assuming what's wrong.
- Prompt fixes. When an issue traces back to agent instructions, you write and test the fix yourself - not just describe the symptom to someone else.
- SQL-driven investigation. Pull your own data from BigQuery/Metabase to confirm or kill a hypothesis, instead of waiting on someone else's report.
- API and webhook debugging. Trace failed webhook deliveries, malformed payloads, and integration issues between client systems and Bolna's platform.
- Evidence-backed escalation. When something genuinely needs an engineer or PM, you hand it off with a documented root cause and reproduction steps - not a forwarded complaint.
- Building the knowledge base. Every new failure mode you diagnose gets written up, so the next person doesn't have to re-diagnose it from scratch.
What you bring
We don't require a specific number of years of experience. We do require that you can walk us through a technical problem you solved by reasoning it out - not by looking up the answer.
- Comfort reading logs, transcripts, and structured data, and forming a real hypothesis about a failure - and revising it when you're wrong.
- Basic SQL - enough to write your own queries to test a hypothesis without waiting on someone else.
- Comfort working with APIs and webhooks - reading a payload, understanding status codes, and figuring out where in a request chain something went wrong.
- Willingness to read and write prompts with precision - understanding why a specific instruction produces a specific model behavior, not just tweaking words until it seems to work.
- Written clarity. A ticket resolution and a bug handoff are both documents. You should enjoy writing them well.
Nice to have
- Any prior exposure to debugging - personal projects, coursework, or hackathons count. We care more about how you reason than where you learned to.
- Knowledge of SQL and know hows of data analysis
- Familiarity with tools like Postman, or having built or consumed a webhook before.
- Prior exposure to voice AI, LLMs, or contact center software.
Success in this role looks like
- Your first hypothesis on a ticket is right more often over time, and when it's wrong, you catch and correct it fast.
- Resolution time comes down for the categories of issues you own.
- Your escalations get accepted without back-and-forth, because the evidence is already there.
- You've documented enough new failure modes that the whole team gets faster at diagnosing that category of problem - not just you.
- Direct access to founders from day one - you’ll be in the room for decisions most people at your stage never see
- Fastest learning curve available in B2B AI sales, product, and customer success
- Real ownership, real customers, real consequences - not a support function
- Value compensation + meaningful ESOP
- Bengaluru office, in-person team collaboration
Skills Required
- Comfort reading logs, transcripts, and structured data and forming hypotheses
- Basic SQL to write queries and investigate data
- Comfort working with APIs and webhooks, understanding payloads and status codes
- Ability to write and refine prompts with precision for LLM behavior
- Strong written communication for ticket resolutions and bug handoffs
- Prior exposure to debugging (projects, coursework, hackathons)
- Knowledge of data analysis practices
- Familiarity with tools like Postman or experience building/consuming webhooks
- Prior exposure to voice AI, LLMs, or contact center software
Am I A Good Fit?
Get Personalized Job Insights.
Our AI-powered fit analysis compares your resume with a job listing so you know if your skills & experience align.
Success! Refresh the page to see how your skills align with this role.
The Company
What We Do
Bolna is a Voice AI Platform purpose-built for India’s scale, linguistic complexity, and cost sensitivity. We enable enterprises to go live with thousands of concurrent calls in days, not months.
