VP of Engineering | Technical Leader | Founding team | Multi-Agent Voice AI | 0→1 Product Builder | IIM | LangGraph · Spring Boot · Java

Hybrid

On-site


Pune Area, India









Chief Product and AI Officer
Indian Institute of Management, Kozhikode

Bachelor of Engineering (B.E.), Computer Science
Pimpri Chinchwad Education Trust'S. Pimpri Chinchwad College Of Engineering
Join Rohan and 10k+ other professionals on
Tal is the app where bosses meet engineers & builders directly, without middlemen
Download Tal BossVP of Engineering | Technical Leader | Founding team | Multi-Agent Voice AI | 0→1 Product Builder | IIM | LangGraph · Spring Boot · Java
I'm a founding engineer turned VP of Engineering, currently building Clara a multi-agent voice AI platform for home services at ZenTrades. Over the last 9+ years I've gone from first-engineer hire to leading platform teams, with a consistent pattern: take a 0→1 idea, architect the foundations, hire the right people, and ship production systems that drive revenue. ➤ What I'm building now (Clara.ai) • Architected and lead engineering for a multi-agent voice AI platform across 6 repositories, Spring Boot 4 Modulith backend, LangGraph Python AI brain, Next.js 16 operator UI, Terraform/EKS infrastructure • Hierarchical LangGraph supervisor routing to 4 specialist agents (answering, scheduling, collections, emergency) over a gRPC bridge • 3-layer auth model Zitadel (authn) + OpenFGA (ReBAC) + Cerbos (ABAC) for clean enterprise multi-tenancy • 38+ ADRs governing identity, voice, billing, and CI/CD • Hire and direct a 6-person team (5 engineers + 1 QA) ➤ Before this (Boomerang / BuyerAssist) • First engineer; built the technical foundation and scaled to a 6-person team • Architected AI-powered B2B SaaS using LangGraph, OpenAI, and modern infra processing millions of contact records daily for enterprise customers • Job Change Detection algorithm at 92%+ precision became the company's primary revenue product • LLM-based lead scoring drove 40%+ improvement in sales efficiency • Salesforce + HubSpot integrations handling millions of sync operations monthly ➤ How I lead • AI-native by default Claude Code, Cursor, Junie, Codex, and Gemini ship production code in my repos under TDD + ADR + automated review gates • Codified standards as rules-as-code so every engineer (and every AI agent) follows the same architectural discipline • Hire for ownership, onboard for context, review for principles not preferences • Decisions become ADRs; ADRs become enforced via CI ➤ What I'm looking for Engineering leadership roles where I can lever 10+ years of 0→1 architecture and team-building. I thrive in founding-team environments, but I'm equally interested in scaling existing organizations. Based in Pune; open to remote globally; US opportunities require visa sponsorship. Currently expanding strategic capabilities via the Chief Product Officer program at IIM Kozhikode (2026). 📩 rohanbhagwat7@gmail.com
Pune Division
claude code
claude
langgraph
langsmith
notebooklm
ollama
claudecode
ChatGPT
VP of Engineering and Platform at Clara AI, Dec 2025 - Present · 8 mos, Hybrid, Founding-team VP leading engineering for Clara, a multi-agent voice AI platform for home services. I own technical strategy, architecture, hiring, and delivery across 6 repos and full-stack infrastructure. ➤ Leadership • Built and lead a 6-person platform team (5 engineers + 1 QA) technical interviews, onboarding, and team permissions • Took Clara from zero to production-ready in under 4 months across 6 repos: Matrix, Operator, Link, Zion, Sentinel, Phosphor • Codified every major decision through 38+ ADRs covering identity, multi-tenancy, voice, billing, and CI/CD • Established AI-native engineering: Claude Code, Cursor, Junie, Codex, and Gemini ship production code under TDD + ADR + CodeRabbit gates ➤ Voice AI • Hierarchical LangGraph supervisor routing to 4 specialist agents (answering, scheduling, collections, emergency) over a gRPC bridge • Real-time voice stack on LiveKit (SIP + WebRTC), Deepgram STT, Cartesia TTS, with SIP REFER warm-transfer and emergency escalation ➤ Backend (Java 25 LTS / Spring Boot 4) • DDD with bounded contexts, event sourcing (outbox), sealed interfaces, immutable records • CRM integration framework with sealed AuthStrategy hierarchy; ServiceTrade and QuickBooks as first-class providers • Airbyte ETL into a canonical customer/invoice/job data model ➤ Identity & Authorization • 3-layer auth: Zitadel OIDC + OpenFGA ReBAC + Cerbos ABAC sidecar clean separation of identity from authorization • Postgres Row-Level Security with workspace-scoped GUC isolation ➤ Infrastructure • EKS 1.34 (Graviton3 ARM64) with Helm releases for Postgres, Valkey, OpenTelemetry, Langfuse, LiveKit, Zitadel, OpenFGA, Cerbos, Airbyte • Terraform-codified VPC, EKS, RDS, WAF, IAM, DNS • Self-hosted GitHub Actions runner with CI for container scan, lint, breaking-change detection ➤ Frontend (Next.js 16 / React 19) • Operator dashboard on @clara/phosphor design system (internal CodeArtifact) • Auth.js v5 with Zitadel OIDC, iron-session, memoized token refresh
Engineer (Founding Team) at Boomerang, May 2021 - Dec 2025 · 4 yrs 8 mos, On-site, Architected and deployed production-grade agentic systems using LangGraph and CopilotKit, building an AI-powered onboarding system that guides users through complex workflows with intelligent copilots. Created a reusable agent framework that abstracted LangGraph complexities into configurable components, enabling rapid development of AI-driven features across the platform. Managed engineering team implementing AI-native development practices where AI coding tools contribute directly to production codebases. Established best practices for AI-assisted development, code review standards for AI-generated code, and quality gates ensuring AI contributions meet production standards. Engineered end-to-end data delivery pipelines processing millions of contact records for enterprise customers, implementing OpenAI-based evaluation systems with custom metrics to automatically assess data relevance and quality. As a founding engineer, drove product development from concept to production (0→1), then established performance management frameworks, code quality standards (CI/CD pipelines, automated testing, code review processes), and team-scaling systems to support enterprise clients (1→N). These practices enabled the team to scale while maintaining velocity. Led development of AI systems deployed in production, including the Job Change (JC) detection algorithm tracking complex job transitions with 92%+ precision. Architected a sophisticated lead tagging framework using LLM-based evaluation to automatically assess and prioritize leads, driving 40%+ improvement in sales efficiency. Designed a scalable IFTTT-style rule engine automating decision-making through dynamic workflows. Engineered enterprise-grade integrations with Salesforce and HubSpot.
Senior Software Engineer at BDO in India, Oct 2020 - May 2021 · 8 mos, Worked on GCOM (Global Compliance and Optimization Management) project for Voyatek: a leading provider of purpose-built software solutions for government and public sector agencies. The project focused on building mission-critical systems for law enforcement and criminal justice organizations. Responsibilities: Fixed bugs in existing criminal record workflows and background check systems Made incremental improvements to justice information sharing features Worked on performance optimization for database queries Maintained CJIS/FBI compliance through security patches and updates Enhanced existing audit logging and access control features Participated in code reviews and testing of changes Adapted to existing architecture and coding standards This role gave me valuable exposure to highly regulated government software and security compliance requirements. Worked in cross-functional and distributed teams across India and US with product managers, security analysts, compliance officers, and government stakeholders.
Founder at RquBits, Nov 2019 - Nov 2020 · 1 yr 1 mo, Founded RquBits as a bootstrapped startup, investing personal savings to build two products from 0→1. Served as technical founder handling product strategy, engineering, and go-to-market while navigating the COVID-19 pandemic. Rooemble - Video Conferencing Platform (April - August 2020): Built Rooemble, a video conferencing application aiming to differentiate in the rapidly growing remote work market. Architected entire platform from scratch using WebRTC for real-time video/audio communication and Jitsi for media server infrastructure. LivelyTicks - Real-Time Stream Analytics (August 2020): Built LivelyTicks, a real-time analytics platform for live streamers to understand audience behavior and engagement. Architected serverless data pipeline processing live stream data at scale. Outcome & Learnings: Shut down operations in September 2020 due to inability to achieve product-market fit and funding constraints. COVID-19 accelerated market competition in video conferencing, making differentiation extremely difficult for bootstrapped venture. This experience taught me invaluable lessons about market timing, customer discovery, burn rate management, and the importance of product-market fit before scaling. The technical and entrepreneurial skills gained directly informed my approach at Boomerang.
Consultant at Self-Employed, Sep 2019 - Oct 2020 · 1 yr 2 mos, Pune Area, India, Provided full-stack engineering, infrastructure, and DevOps consulting services to clients, Taking complete ownership of critical technical initiatives from architecture to deployment. Sapper SDK - Integration Platform: Co-designed and implemented core SDK functionality for SaaS integration platform connecting disparate business applications. Architected API design patterns, OAuth authentication mechanisms, webhook handling, and integration connectors. Built robust error handling and retry logic for reliable third-party integrations. Distributed Job Scheduling - Quartz Clustering: Took complete ownership of implementing Quartz Clustering for distributed job scheduling across multiple servers. Designed high-availability architecture ensuring fault tolerance for critical background jobs. Implemented job persistence, coordination across nodes, and automatic failover. This solution enabled clients to scale horizontally while maintaining reliable scheduled task execution. High-Performance Caching Architecture: Architected comprehensive Redis-based caching layer that improved application performance by 10x. Designed cache invalidation strategies, implemented distributed locking for consistency, and optimized cache hit rates through intelligent key design. Reduced database load by 80%+ and improved user-facing API response times from seconds to milliseconds. DevOps & CI/CD Infrastructure: Took complete ownership of establishing CI/CD pipelines using Jenkins. Implemented automated testing frameworks (unit, integration, end-to-end), containerization with Docker, deployment automation, and infrastructure monitoring. Created blue-green deployment strategies for zero-downtime releases. Reduced deployment time from hours to minutes while improving reliability.
Chief Product and AI Officer, Indian Institute of Management, Kozhikode
Bachelor of Engineering (B.E.), Computer Science, Pimpri Chinchwad Education Trust'S. Pimpri Chinchwad College Of Engineering