Apul Gupta profile photo 1 on Tal

Apul Gupta

AI Product Leader | GenAI & Agentic AI | LangGraph · RAG · LLM Governance | HSBC · Google · Amazon · EY

Bengaluru

7+ years of work experience

HSBC
Sep 20262 mosPresent

Leading GenAI product strategy and platform delivery at HSBC across India, UAE, UK, Australia & Singapore. Managing 14 GenAI engineers across 6+ cross-functional squads and 4 concurrent AI programmes with multi-million-dollar technology budgets. GenAI Agent Assist – Led architecture and delivery of a production-grade multi-agent platform for Collections & Fraud using LangGraph, RAG, real-time transcription and enterprise knowledge integration. Delivered 30% reduction in AHT and $1.23M annual savings, supporting thousands of customer interactions daily. GenAI Origination & Underwriting – Leading development of an agentic AI platform for credit and mortgage origination using multimodal LLMs and specialised agents for document classification, extraction, validation, fraud checks, policy evaluation, affordability and credit memo generation. Targeting 50% faster processing and $1M+ annual savings, enabling same-day origination journeys across UAE & Singapore. Responsible AI & LLM Governance – Driving GenAI production readiness across Legal, Resilience, Cybersecurity, Financial Crime, People & Model Risk. Partnering with global/regional risk and regulatory teams on LLM evaluation, monitoring and controls for responsible AI deployment. AI Economics & Optimisation – Driving AI cost modelling, capacity forecasting and cost-benefit analysis across LLM, speech, infrastructure, token and compute costs to build scalable AI business cases. Fraud Intelligence Management – Leading an AI-powered fraud investigation platform for HSBC UAE combining anomaly detection, risk scoring, cross-channel intelligence and AI-generated investigation reports, targeting ~60% reduction in investigation time.
HSBC
Oct 20251 yr 1 moPresent

HSBC
Dec 202411 mosOct 2025

Google
Aug 20229 mosApr 2023

Bengaluru, Karnataka, India

Technical Program Manager for Google Cloud Spanner, owning enterprise product launches, API lifecycle strategy, and developer ecosystem growth across 100+ enterprise clients globally. Key deliveries: Fine-Grained Access Control Launch - Owned end-to-end launch of IAM-integrated SQL-level RBAC across 10+ enterprise clients, strengthening audit compliance, enabling granular role management at scale, and reducing unauthorised access risk across regulated industries. Dual-Region APRA Compliance - Delivered dual-region Cloud Spanner configuration across Australian zones with synchronous replication and cross-zone failover to satisfy APRA data-sovereignty mandates, improving latency performance, fault tolerance, and regional SLA compliance for financial services clients. SDK & API Lifecycle Ownership - Owned OKRs and product roadmap for Spanner client libraries and SDKs, leading a distributed team of 45+ software engineers in collaboration with Product, SRE, and DevRel teams across 100+ enterprise clients. Drove end-to-end API lifecycle management spanning authentication, versioning, SDK delivery, and developer experience. Global Engineering Standardisation - Architected and drove adoption of a shared development model across global partner teams, standardising engineering processes and boosting launch velocity by 20% across 10+ internal functions and 12+ regional stakeholders. Developer Support Excellence - Built comprehensive support SLA and escalation framework for GitHub and Stack Overflow developer channels, leading structured root cause analysis initiatives that reduced average developer turnaround time by 30%.
Amazon
Nov 20201 yr 9 mosJul 2022

Hyderabad, Telangana, India

Program Manager for Last Mile Analytics at Amazon, leading AI-driven fraud prevention, platform security, and operational analytics across the US, UK, Singapore, and India. Key deliveries: Bot Detection & GPS Spoofing Prevention - Led cross-functional rollout of real-time fraud detection for Amazon Flex using telemetry analytics, behavioural modelling, and anomaly-detection ML models. Reduced third-party fraud by 60% and protected platform integrity across thousands of daily delivery transactions. Driver Risk Scoring Engine - Designed and deployed an ML-powered driver risk-scoring system using GPS drift, delivery-time variance, and behavioural-anomaly signals integrated with a decision-tree model. Reduced manual fraud investigations by 40% and streamlined fraud triage operations at scale. AWS Rekognition OCR Pipeline - Implemented computer vision-based driver licence verification pipeline using AWS Rekognition, detecting duplicate profiles and reducing onboarding fraud by 70% across the Flex driver network. Global Programme Scaling - Scaled fraud-prevention programmes across US, UK, and Singapore by establishing governance frameworks, regional coordination models, and standardised playbooks that sustained $10M+ in cost efficiencies across markets. Executive Analytics & Performance Intelligence - Built gamified Tableau performance dashboard for real-time auditor tracking, driving 25% productivity gains and embedding data-driven insights into VP and Director-level MBRs for strategic decision-making. Driver Compliance & Experience - Launched PAN-based identity verification for Indian drivers improving regulatory compliance for 35%+ of new enrolments. Partnered with L&D to design cross-skilling programmes combining interactive training and performance dashboards, achieving 20% NPS improvement.
EY
Apr 20182 yrs 8 mosNov 2020

Bengaluru, Karnataka, India

Associate Consultant in IT Advisory at Ernst and Young, delivering digital platform scaling, AI-driven analytics, and government infrastructure advisory across national-scale programmes in India. Key deliveries: DIKSHA National Education Platform - Drove 250x growth in app usage for India's national education platform by architecting real-time analytics dashboards with vernacular language support, enabling 100+ education officials across 15+ states to deliver data-driven outreach at national scale. One of the largest government digital platform scaling engagements in India at the time. Predictive Traffic Forecasting - Built ARIMA-based user-traffic and content-access forecasting models in collaboration with DevOps and cloud engineering teams, maintaining 99.99% system availability and achieving cost-efficient AWS infrastructure scaling across peak demand periods. Python Automation & Workflow Intelligence - Automated content tracking and reporting workflows using Python, reducing manual effort by 60% and saving 20+ analyst hours per week across 5,000+ QR-linked digital textbooks, freeing the team to focus on higher-value analytical work. $480M State Digital Infrastructure Programme - Designed optical-fibre network architecture for 10,000+ villages defining FTTH topology, PoP locations, and core-backhaul design to expand broadband connectivity across underserved regions under one of India's largest state infrastructure investments. Government Vendor Advisory - Led end-to-end vendor evaluation and onboarding for 15+ system integrators, collaborating with state government and technical advisory bodies to assess broadband demand, draft RFP specifications, and ensure timely programme delivery against regulatory milestones.
Wipro Limited
Oct 20177 mosApr 2018

New Delhi, Delhi, India

Network Operations Center Administrator at Wipro, building ML-powered security and monitoring infrastructure across a 1,000+ endpoint enterprise network. This role established the engineering and ML foundation that underpins 8 years of AI product delivery at Amazon, Google, and HSBC. Key deliveries: ML-Powered Network Anomaly Detection - Built and deployed a machine learning-based threat detection system using Scikit-learn and regex log parsers, providing real-time monitoring across 1,000+ endpoints with automated detection of port scans, brute-force attacks, and unauthorised access attempts. Improved incident response time by 40%. CI/CD & MLOps Pipeline Automation - Automated end-to-end monitoring and ML pipeline deployment using Linux, Git, cron jobs, Jenkins, and ELK Stack, implementing CI/CD workflows that boosted operational uptime by 20%.
GitHub
@apul1421

Apul’s AI Stack

anthropic apis
anthropic apis
antigravity
antigravity
codex
codex
github copilot
github copilot
hugging face
hugging face
langchain
langchain

Education

Liverpool John Moores University

Master's degree, Artificial Intelligence and Machine Learning

Liverpool John Moores University

20211 yr2022
International Institute of Information Technology Bangalore

Post Graduate Diploma in Machine learning and Artificial Intelligence, Computer Science

International Institute of Information Technology Bangalore

20191 yr2020

Join Apul and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Apul Gupta on Tal

Headline

AI Product Leader | GenAI & Agentic AI | LangGraph · RAG · LLM Governance | HSBC · Google · Amazon · EY

About

AI Product Leader, 9+ years across HSBC, Google, Amazon, and EY. Shipped GenAI products worth $50M+ in business impact in credit, contact centre, and financial crime. I combine hands-on LLM and agentic AI depth with the stakeholder work needed to get AI live in regulated environments. If you are building something serious with AI and need someone who has shipped it in regulated environments at global scale, let us connect.

Location

Bengaluru

AI stack

anthropic apis

antigravity

codex

github copilot

hugging face

langchain

Work experience

Associate Director - Innovation (GenAI and Agentic AI) at HSBC, Sep 2026 - Present · 2 mos, Leading GenAI product strategy and platform delivery at HSBC across India, UAE, UK, Australia & Singapore. Managing 14 GenAI engineers across 6+ cross-functional squads and 4 concurrent AI programmes with multi-million-dollar technology budgets. GenAI Agent Assist – Led architecture and delivery of a production-grade multi-agent platform for Collections & Fraud using LangGraph, RAG, real-time transcription and enterprise knowledge integration. Delivered 30% reduction in AHT and $1.23M annual savings, supporting thousands of customer interactions daily. GenAI Origination & Underwriting – Leading development of an agentic AI platform for credit and mortgage origination using multimodal LLMs and specialised agents for document classification, extraction, validation, fraud checks, policy evaluation, affordability and credit memo generation. Targeting 50% faster processing and $1M+ annual savings, enabling same-day origination journeys across UAE & Singapore. Responsible AI & LLM Governance – Driving GenAI production readiness across Legal, Resilience, Cybersecurity, Financial Crime, People & Model Risk. Partnering with global/regional risk and regulatory teams on LLM evaluation, monitoring and controls for responsible AI deployment. AI Economics & Optimisation – Driving AI cost modelling, capacity forecasting and cost-benefit analysis across LLM, speech, infrastructure, token and compute costs to build scalable AI business cases. Fraud Intelligence Management – Leading an AI-powered fraud investigation platform for HSBC UAE combining anomaly detection, risk scoring, cross-channel intelligence and AI-generated investigation reports, targeting ~60% reduction in investigation time.

Assistant Director - Innovation (GenAI and Agentic AI) at HSBC, Oct 2025 - Present · 1 yr 1 mo

Lead Assistant Vice President - Innovation (GenAI and Agentic AI) at HSBC, Dec 2024 - Oct 2025 · 11 mos

Technical Program Manager at Google, Aug 2022 - Apr 2023 · 9 mos, Bengaluru, Karnataka, India, Technical Program Manager for Google Cloud Spanner, owning enterprise product launches, API lifecycle strategy, and developer ecosystem growth across 100+ enterprise clients globally. Key deliveries: Fine-Grained Access Control Launch - Owned end-to-end launch of IAM-integrated SQL-level RBAC across 10+ enterprise clients, strengthening audit compliance, enabling granular role management at scale, and reducing unauthorised access risk across regulated industries. Dual-Region APRA Compliance - Delivered dual-region Cloud Spanner configuration across Australian zones with synchronous replication and cross-zone failover to satisfy APRA data-sovereignty mandates, improving latency performance, fault tolerance, and regional SLA compliance for financial services clients. SDK & API Lifecycle Ownership - Owned OKRs and product roadmap for Spanner client libraries and SDKs, leading a distributed team of 45+ software engineers in collaboration with Product, SRE, and DevRel teams across 100+ enterprise clients. Drove end-to-end API lifecycle management spanning authentication, versioning, SDK delivery, and developer experience. Global Engineering Standardisation - Architected and drove adoption of a shared development model across global partner teams, standardising engineering processes and boosting launch velocity by 20% across 10+ internal functions and 12+ regional stakeholders. Developer Support Excellence - Built comprehensive support SLA and escalation framework for GitHub and Stack Overflow developer channels, leading structured root cause analysis initiatives that reduced average developer turnaround time by 30%.

Program Manager at Amazon, Nov 2020 - Jul 2022 · 1 yr 9 mos, Hyderabad, Telangana, India, Program Manager for Last Mile Analytics at Amazon, leading AI-driven fraud prevention, platform security, and operational analytics across the US, UK, Singapore, and India. Key deliveries: Bot Detection & GPS Spoofing Prevention - Led cross-functional rollout of real-time fraud detection for Amazon Flex using telemetry analytics, behavioural modelling, and anomaly-detection ML models. Reduced third-party fraud by 60% and protected platform integrity across thousands of daily delivery transactions. Driver Risk Scoring Engine - Designed and deployed an ML-powered driver risk-scoring system using GPS drift, delivery-time variance, and behavioural-anomaly signals integrated with a decision-tree model. Reduced manual fraud investigations by 40% and streamlined fraud triage operations at scale. AWS Rekognition OCR Pipeline - Implemented computer vision-based driver licence verification pipeline using AWS Rekognition, detecting duplicate profiles and reducing onboarding fraud by 70% across the Flex driver network. Global Programme Scaling - Scaled fraud-prevention programmes across US, UK, and Singapore by establishing governance frameworks, regional coordination models, and standardised playbooks that sustained $10M+ in cost efficiencies across markets. Executive Analytics & Performance Intelligence - Built gamified Tableau performance dashboard for real-time auditor tracking, driving 25% productivity gains and embedding data-driven insights into VP and Director-level MBRs for strategic decision-making. Driver Compliance & Experience - Launched PAN-based identity verification for Indian drivers improving regulatory compliance for 35%+ of new enrolments. Partnered with L&D to design cross-skilling programmes combining interactive training and performance dashboards, achieving 20% NPS improvement.

Associate Consultant at EY, Apr 2018 - Nov 2020 · 2 yrs 8 mos, Bengaluru, Karnataka, India, Associate Consultant in IT Advisory at Ernst and Young, delivering digital platform scaling, AI-driven analytics, and government infrastructure advisory across national-scale programmes in India. Key deliveries: DIKSHA National Education Platform - Drove 250x growth in app usage for India's national education platform by architecting real-time analytics dashboards with vernacular language support, enabling 100+ education officials across 15+ states to deliver data-driven outreach at national scale. One of the largest government digital platform scaling engagements in India at the time. Predictive Traffic Forecasting - Built ARIMA-based user-traffic and content-access forecasting models in collaboration with DevOps and cloud engineering teams, maintaining 99.99% system availability and achieving cost-efficient AWS infrastructure scaling across peak demand periods. Python Automation & Workflow Intelligence - Automated content tracking and reporting workflows using Python, reducing manual effort by 60% and saving 20+ analyst hours per week across 5,000+ QR-linked digital textbooks, freeing the team to focus on higher-value analytical work. $480M State Digital Infrastructure Programme - Designed optical-fibre network architecture for 10,000+ villages defining FTTH topology, PoP locations, and core-backhaul design to expand broadband connectivity across underserved regions under one of India's largest state infrastructure investments. Government Vendor Advisory - Led end-to-end vendor evaluation and onboarding for 15+ system integrators, collaborating with state government and technical advisory bodies to assess broadband demand, draft RFP specifications, and ensure timely programme delivery against regulatory milestones.

Software Engineer at Wipro Limited, Oct 2017 - Apr 2018 · 7 mos, New Delhi, Delhi, India, Network Operations Center Administrator at Wipro, building ML-powered security and monitoring infrastructure across a 1,000+ endpoint enterprise network. This role established the engineering and ML foundation that underpins 8 years of AI product delivery at Amazon, Google, and HSBC. Key deliveries: ML-Powered Network Anomaly Detection - Built and deployed a machine learning-based threat detection system using Scikit-learn and regex log parsers, providing real-time monitoring across 1,000+ endpoints with automated detection of port scans, brute-force attacks, and unauthorised access attempts. Improved incident response time by 40%. CI/CD & MLOps Pipeline Automation - Automated end-to-end monitoring and ML pipeline deployment using Linux, Git, cron jobs, Jenkins, and ELK Stack, implementing CI/CD workflows that boosted operational uptime by 20%.

Education

Master's degree, Artificial Intelligence and Machine Learning, Liverpool John Moores University

Post Graduate Diploma in Machine learning and Artificial Intelligence, Computer Science, International Institute of Information Technology Bangalore

Public links

github: https://github.com/apul1421