Software Engineer
FloBiz · MyCashFlo · Remote (India) · Full-time
About FloBiz
FloBiz builds AI for the world’s small businesses. Our flagship product, myBillBook, is India’s #1 billing, stock management & accounting platform, trusted by over 1 crore MSMEs across 18,000+ pincodes. On top of that foundation, we are now building a family of agentic AI products. This is AI that doesn’t just answer questions but does the work: making calls, booking appointments, collecting payments, and closing loops that used to need a full-time employee.
Today that family of products live in two markets: MyCashFlo, our AI collections manager for Indian businesses, and heyAnaya, our AI front desk for clinics in the UAE. Both run on a voice AI platform we built entirely in-house, with 881,000+ completed calls and counting across 12 languages from Hindi to Gulf Arabic.
We are backed by Peak XV Partners, Elevation Capital, Greenoaks, BEENEXT and Think Investments. Our tech and product teams are remote by design, with in-person offsites, hackathons and team weeks to bring everyone together.
About MyCashFlo
MyCashFlo is an AI-first product built for one of the hardest tests of AI: a live phone call, in an Indian language, with someone who would rather not be on it. We place over 30,000 collection calls a day, and that number is climbing fast.
Every one of those calls is a complete LLM loop running inside a sub-second latency budget - speech recognition, reasoning, tool invocation, speech synthesis across Indian accents, code-switched Hindi-English, unpredictable ambient noise, and a dozen languages most ASR models were never trained to hear properly. Our agents carry memory across conversations, so a third call opens knowing exactly what was committed on the second. They invoke tools mid-sentence: pushing UPI details to WhatsApp the moment a customer asks how to pay, or warm-transferring the live call to the business owner when the conversation genuinely needs a human.
The entire voice AI stack behind the call is built in-house. Multi-vendor telephony routing and caller-ID pool management. Prompt and persona architectures versioned, tested and tuned like production systems, with the eval infrastructure to back them. A post-call intelligence layer that converts unstructured speech into structured outcomes - intent classification with confidence scoring, commitment extraction, sentiment, next action. And an orchestration layer that keeps thousands of concurrent conversations reliable, attributable and un-repeatedly metered.
The problems worth solving here are still open. How do you make an agent sound like a person rather than a script? How do you evaluate a system whose output is a conversation? How do you keep a model honest when it mishears a date and confidently manufactures a promise? And how do you take all of it from a demo that impresses to a fleet that runs every single day, for real businesses, with real money on the line?
If you want to build agents that operate in the physical world rather than a chat window, this is the team.
What You’ll Do
- Work in a fast-paced, agile culture shipping features with immediate business impact. Your code changes how a real business collects money or books its patients, often within days of merging.
- Own features end to end across the platform, collaborating closely with product, frontend and DevOps teams.
- Build and scale the orchestration systems that keep thousands of concurrent AI conversations reliable: call scheduling, multi-vendor telephony routing, wallets and metering, and post-call processing pipelines where every rupee and every commitment must be accounted for exactly once.
- Work on the post-call intelligence layer that turns unstructured conversations into structured, actionable outcomes: intents, commitments, sentiment and next actions.
- Strengthen the evals and observability infrastructure that keeps a fleet of AI agents honest in production.
- Connect the agents to the real world: payments and WhatsApp, calendars and EMR systems, telephony and messaging vendors.
- Proactively identify technical initiatives that strengthen the scalability, security and maintainability of the platform.
- Write highly scalable and maintainable code while keeping agility in mind.
- Build with AI, not just on it. AI-powered development tools and workflows are how we work every day, and we expect you to push that frontier with us.
What We’re Looking For
- 3+ years of core software engineering experience, preferably in a product startup.
- Hands-on development experience with at least 2 programming languages, including dynamically typed and object-oriented languages, preferably Java, Python and Ruby along with their corresponding web development frameworks.
- Working experience with SQL and NoSQL databases and their advanced concepts.
- Ability to write code and APIs with DRY/KISS and SOLID principles in mind.
- Experience writing unit and integration tests with good code coverage.
- Hands-on experience designing and working with microservices.
- Experience working on scalable systems handling a high volume of requests
- Curiosity about how LLM-powered systems behave in production. Exposure to LLM APIs, agent frameworks, prompt engineering or eval tooling is a strong plus.
- A proactive approach to leveraging AI in software development, with an eagerness to explore and implement AI-assisted workflows for enhanced productivity and problem-solving.
Perks & Benefits
- Extensive medical insurance that looks out for our employees and their dependents. We’ll love you and take care of you, our promise.
- FloBiz Academy to help you keep learning and enhancing your skills.
- Competitive salary, with a reward system that celebrates hard work, milestones and performance throughout the year.
- Remote by design: work from anywhere in India, with in-person offsites, hackathons and team weeks that bring the whole team together.
Location: Remote (India)