Work Experience
Mastercard Pune, India
Director – Product Development, Agentic AI May 2026 – Present
- Architecting agentic AI systems at Mastercard scale for operational reports and intelligence workflows.
Layerpath [a16z Speedrun'25 Backed Startup] Remote – Coimbatore, India
Chief Technology Officer (CTO) Mar 2025 – Apr 2026
- Architected "Path AI", a voice-first inbound lead qualification system handling 1,500+ calls/month with 85%+ accuracy and 75% cost reduction vs. human SDRs.
- Designed 7-stage FSM (BANT and Four Forces) with hybrid rule/LLM state transitions and a ReAct reasoning loop validating responses against RAG knowledge to prevent hallucinations.
- Cut voice latency below 1s via concurrent STT/LLM/TTS streaming with Pipecat, Daily.co WebRTC, Deepgram, and Cartesia/ElevenLabs.
- Built tool orchestration layer to trigger calendar booking, CRM updates, Slack alerts, and webhook workflows from live conversation context.
- Implemented guardrails for PII redaction, topic drift detection, and prompt injection alongside Langfuse tracing, per-call cost attribution, and synthetic eval suites.
- Scaled product from zero to $5K MRR; led 2 engineers and 1 designer across hiring, architecture, and deployment.
Cloud Ambassadors [Google Cloud Partner] Remote – Coimbatore, India
Director of Engineering | Data & AI Jul 2024 – Feb 2025
- Led 40 engineers across 3 teams delivering production AI agents, document intelligence platforms, and RAG search for Google Cloud enterprise clients.
- Architected multi-tenant hybrid search platform (vector + BM25 fusion) indexing 1M+ documents on GCP with sub-second retrieval latency.
- Built document intelligence pipeline processing 100K+ files/month at 90%+ extraction accuracy across PDFs, images, and handwriting via Vertex AI vision models.
- Created internal agentic framework with reusable patterns for tool orchestration, prompt chains, and RAG retrieval across client projects.
- Cut GCP GenAI infrastructure costs through semantic caching, model routing, and batch embedding pipelines.
Tookitaki [Thunes Holdings] Singapore
Engineering Manager | Data & Full Stack Sep 2022 – May 2024
- Managed 24 engineers across 3 units, delivering batch and streaming AML and anti-fraud platforms processing 10M+ transactions daily across multi-region deployments.
- Redesigned anti-fraud engine using Apache Flink, increasing throughput 19x with sub-100ms real-time transaction scoring.
- Built ML model serving infrastructure and a configurable rule engine, allowing compliance teams to update fraud detection logic without engineering releases.
- Cut annual infrastructure OPEX by 30% by replacing legacy Hadoop and HBase clusters with ScyllaDB and Spark on Kubernetes.
Agoda [Booking Holdings Inc.] Bangkok, Thailand
Staff Data Engineer | Data Platform Team May 2022 – Aug 2022
- Led petabyte-scale storage migration from HDFS to Vast S3-compatible storage, supporting 150+ TB/day ingestion.
- Built centralized credential broker using HashiCorp Vault, securing secrets across 1,000+ production data pipelines.
Senior Data Engineer | Data Platform Team Aug 2021 – May 2022
- Prototyped lakehouse architecture using Apache Iceberg for ACID transactions, alongside CDC pipelines via Debezium and Apache Flink.
- Created self-service platform tools adopted by 190+ data teams; led code reviews and mentored junior engineers.
Data Engineer | Data Platform Team Jan 2020 – Aug 2021
- Built self-service ETL platform processing 150+ TB/day across 1,000+ pipelines, with automated inter-cluster migration tools.
Senior Software Dev. Engineer in Test (SDET) | Data Platform Team Jan 2019 – Dec 2019
- Built automated data validation framework using Spark and Koalas with schema drift detection across Big Data pipelines.
- Created pipeline performance benchmarking suite integrated into CI/CD to catch throughput regressions before deployment.
Early Career Experience India Test Automation & Quality Engineering Roles Jul 2012 – Dec 2018
Show details ↓
Lead/Senior Software Dev. Engineer in Test (SDET) | redBus [MakeMyTrip - Goibibo Group] Mar 2017 – Dec 2018
- Built core API test automation framework from scratch and integrated it into CI/CD for release validation.
- Conducted high-volume load testing on inventory booking systems to validate stability under holiday traffic peaks.
- Defined automation standards and collaborated with product teams on quality requirements.
Test Automation Engineer | Scientific Games Corporation (Atlantic Lottery) Apr 2016 – Mar 2017
- Ran load and stress testing on lottery APIs handling millions of transactions, tuning microservices to maintain sub-100ms draw responses.
- Identified and resolved microservices bottlenecks across high-throughput lottery platforms.
Test Automation Engineer | Cognizant [Client: LexisNexis] Jul 2012 – Apr 2016
- Built test automation framework covering 100+ RESTful endpoints for the Lexis Advance legal search platform.
- Developed search relevance testing suite validating document retrieval accuracy and ranking algorithms.
Skills
Agentic AI & GenAI: Pipecat, Daily.co WebRTC, Deepgram, Cartesia, ElevenLabs, Langfuse, DSPy, BAML, Milvus, Vertex AI, Gemini, OpenAI & Anthropic APIs
Programming Languages: Python, Scala, Rust, Java, TypeScript, SQL
Frameworks & DBs: FastAPI, React, Next.js, Node.js, Postgres, Supabase, ScyllaDB, Redis, Debezium
Big Data Technologies: Apache Spark, Apache Flink, Kafka, Apache Iceberg, Airflow, HBase, Hive, BigQuery, Redshift
Cloud Platforms: Google Cloud (GKE, Cloud Run, Vertex AI, BigQuery), AWS (S3, Lambda, EC2, Bedrock)
DevOps & Infra: Kubernetes, Docker, Terraform, GitHub Actions, HashiCorp Vault, Prometheus, ELK Stack
Certifications
Google Cloud – Professional Data Engineer ↗ Feb 2025 – Feb 2027
Education
PSG College of Technology | Anna University Coimbatore, India
B.Sc Information Technology
CGPA: 9.43/10 Jun 2009 – May 2012