Director of Product · AI Architect · India

Archisman
builds things
that scale.

14+ years building consumer apps, enterprise SaaS, fintech infrastructure, and production AI systems -- from concept to global scale. Pioneering automation and AI-driven transformation across industries -- turning complex problems into intelligent, scalable products.

Edvoy — 1M+ users · 90+ countries
Colive — 550% revenue growth
AbsolutPrep — Exit achieved
Impact at a glance
1M+
Users globally
550%
Revenue growth
86%
Retention uplift
8
AI systems live
Director of Product · Edvoy
🧠
8 AI Systems
RAG · Agents · LLMs · GenAI
🌎
90+ Countries
Global product scale
🚀
0 → 1 → 100
Full build lifecycle
Scroll
Industries Impacted
Broad range.
Deep roots.
Products that solve real problems don't respect industry boundaries. Built, shipped, and scaled across ten distinct domains.
10
01
🎓
Edtech
1M+ students · 90+ countries
02
🏠
Proptech
550% revenue growth platform
03
💳
Fintech
Payments, lending & cross-border
04
🛒
E-commerce
AI-powered commerce & OMS
05
📡
Telco
Fortune 50 enterprise systems
06
👥
HR Tech
End-to-end HR SaaS platform
07
📦
Supply Chain
Intelligent automation
08
🚚
Logistics
Fulfilment & order management
09
🏥
Healthcare
AI-assisted clinical workflows
10
🤝
Consulting
Product & AI transformation
01 / Who I Am

Product mind.
AI instinct.

I'm a Director of Product who has spent 14 years building at the intersection of complex systems, user behaviour, and emerging technology. My work spans 0-to-1 consumer apps, ERP architecture from scratch, multi-sided platforms, fintech infrastructure, and production AI systems -- not as experiments, but as businesses that ship and grow.

At Edvoy, I own four product verticals simultaneously: a consumer app (1M+ users, 90+ countries), Edvoy Edge (enterprise SaaS), Edvoy Cloud (partner interface), and an internal ERP with 35+ modules. I have shipped four AI systems into production -- RAG knowledge engines, agentic pipelines, LLM chatbots, and GenAI content platforms.

Before Edvoy: Head of Product at Colive (Sattva Group) where I built a 32-module ERP from zero contributing to 550% revenue growth, AVP Products at Univariety (Info Edge), and co-founder plus exit at AbsolutPrep. Earlier career spans Paytm, Lodestar, and Verizon.

Beyond core product work, I take on consulting engagements across e-commerce, home decor, retail, and digital transformation -- building products end-to-end including AI systems, commerce platforms, and SaaS tools for clients across industries.

Consumer AppsEnterprise SaaSAI / Agentic FintechEdtechPLG 0 to 1 BuilderERP ArchitectureObservability
6
Product Verticals Owned
35+
ERP Modules Built from Scratch
6500+
Businesses Served on Enterprise SaaS
0→1→100
At Global Scale
86%
Retention Rate Increase
14+
Years of Product Leadership
5+
New Revenue Streams Built for Edvoy
90+
Countries on a Single Platform
02 / Career

Fourteen years.
Constant growth.

2022 -- PresentCurrent
Director of Product
Edvoy -- Global Education Platform · Hyderabad, India
Own four product verticals: consumer app (1M+ users, 90+ countries), Edvoy Edge (enterprise SaaS), Edvoy Cloud (partner interface), and a 35+ module internal ERP. Disbursed 24M GBP in student funding across 20+ banking partners. Shipped four production AI systems -- RAG, agentic pipeline, LLM chatbot, GenAI content engine. 750+ institutional partners across the platform.
Consumer AppEnterprise SaaSRAG GenieAgentic PipelineLLM ChatbotGenAIEdFintech
2020 -- 2022
AVP Products
Univariety (Info Edge) · Hyderabad, India
Led product for India's leading higher-education guidance platform. Drove college-fit personalisation algorithms, alumni mentorship networks, and institutional SaaS counselling tooling. Scaled the consumer platform to serve lakhs of students and thousands of institutional counsellors across India.
Consumer PlatformPersonalisationInstitutional SaaSCollege Fit AI
2019 -- 2020
Head of Product
Colive (Sattva Group) · Bangalore, India
Built a 32-module ERP from the ground up -- property management, billing, resident experience, ops automation -- directly driving 550% revenue growth. Designed a white-label SaaS platform driving an 86% retention rate increase. Built the resident-facing consumer app. Led PLG strategy across the full product surface.
0-1 ERPConsumer AppWhite-label SaaSPLG550% Revenue
2017 -- 2019
Product Manager
Lodestar · Bangalore, India
Built media intelligence and analytics products at one of India's largest media agencies. Developed data dashboards, campaign performance tools, and audience targeting systems across FMCG, BFSI, and consumer durables clients.
Media IntelligenceAnalyticsCampaign ToolsBFSI
2016 -- 2017
Co-founder
AbsolutPrep · Delhi, India
Co-founded and built an ed-tech test-prep platform from zero to exit. End-to-end ownership across adaptive assessments, content management, and student analytics. Successful exit within 12 months of founding.
FoundingSuccessful ExitAdaptive LearningConsumer App
2011 -- 2015
Business Analyst to Product
Verizon · Hyderabad, India
Started career at Verizon building enterprise telco product and analytics capabilities. Built deep foundations in large-scale systems, enterprise workflows, data pipelines, and stakeholder management within a Fortune 50 environment.
TelcoEnterprise SystemsData PipelinesFortune 50
Career Arc
Edvoy
'22
Univariety
'20
Colive
'19
Lodestar
'17
AbsolutPrep
'16
Verizon
'11
Career Highlights
14+
Years in product
6
Companies
1
Co-founded & exited
10+
Industries impacted
Domain Evolution
AI & Machine Learning
Enterprise SaaS
0 to 1 Building
Consumer Products
Fintech
Platform & ERP
HR Tech
Consulting
Logistics
E-commerce
Education & Certifications
🏫
MBA
Great Lakes Institute of Management
📊
International Business
IIM Bangalore
🏫
Marketing Analytics
University of Illinois
📚
B.Tech
Techno India
SAFe PO/PM Certified
Scaled Agile
03 / AI Systems

Not demos.
End-to-end builds.

Every system below was conceived, designed, built, and launched entirely by me -- covering the full lifecycle: problem statement -- strategy -- product design -- content -- architecture -- coding -- environment setup -- deployment -- GTM -- outcomes.

01
RAG Knowledge Engine · Edvoy Production
Genie -- AI Knowledge Engine
Problem Statement
Students exploring study-abroad options had to navigate hundreds of university pages, visa policy docs, and scholarship portals to get answers. Counsellor queues were overloaded and response times were too slow for a global, always-on student base across 90+ countries.
A production RAG system powering intelligent Q&A across Edvoy's entire knowledge base -- universities, courses, visa, scholarships, FAQs -- serving 1M+ students with sub-second responses and hallucination guardrails.
🎯
Planning & Strategy
  • Defined RAG architecture vision for multi-domain queries across 90+ countries
  • Mapped knowledge domains: 750+ institutions, visa policies, scholarships, FAQs
  • Prioritised semantic vs fixed-size chunking strategy per content type
  • Aligned on latency SLAs, accuracy thresholds, and hallucination guardrails
🎨
Product Design & UX
  • Conversational Q&A interface with source attribution and confidence signals
  • Fallback flows for low-confidence retrievals with escalation to counsellors
  • Thumbs up/down feedback loop to continuously retrain retriever quality
  • Admin panel for knowledge ingestion, refresh scheduling, and monitoring
📝
Content & Knowledge Base
  • Audited and structured 750+ institutional data sources with metadata schema
  • Built content refresh pipelines for live data sync with institutional feeds
  • Wrote system prompts, Genie persona, and tone guidelines
  • Created QA test sets per domain to benchmark retrieval accuracy
Technical Build & Environments
  • Vector DB (Pinecone/Weaviate), embedding pipeline, chunking, re-ranking layer
  • LLM integration with context injection and guardrails against hallucination
  • Docker containerisation, CI/CD, staging to production environment promotion
  • Latency optimisation: caching layer, async retrieval, batch embedding jobs
🚀
Deployment & GTM
  • Phased rollout: internal QA -- counsellor beta -- student-facing production
  • KPIs defined: query resolution rate, CSAT, support ticket deflection %
  • In-app onboarding with seeded example questions to drive adoption
  • Post-launch conversation quality audits and weekly prompt iteration
Support ticket volume reducedSub-second Q&A at 1M+ scale90+ countries coveredAlways-on student support
RAGVector DBSemantic SearchRe-rankingLLMDockerCI/CD
02
Agentic System · Edvoy Production
Student Nurture Agent Pipeline
Problem Statement
With 1M+ students across 90 countries, manual counsellor-driven nurture was unscalable. Students dropped off the funnel due to delayed, generic follow-ups -- not the right content at the right time.
Autonomous multi-agent pipeline orchestrating personalised student engagement from interest signal to application intent -- profiling, content sequencing, nudge timing, and counsellor handoff.
🎯
Planning & Agent Design
  • Mapped full student funnel: discovery -- shortlist -- application -- enrolment
  • Identified 12 trigger events for agent activation across the funnel stages
  • Designed agent roles: Profiler, Content Agent, Nudge Agent, Handoff Agent
  • Defined human-in-the-loop thresholds for counsellor escalation
Build, Deploy & GTM
  • LangGraph-style orchestration with tool use and per-student memory persistence
  • Integrated across CRM, email, push notifications, and WhatsApp channels
  • A/B tested nudge cadence, content format, and timing per segment
  • Event-driven architecture with real-time monitoring and intervention dashboard
Funnel conversion improved~40% reduction in manual counsellor workload
Agentic AILangGraphTool UseCRM Integration
🤖
03
Conversational AI · Edvoy Production
LLM Chatbot
Problem Statement
Students needed immediate, contextual answers during off-hours when counsellors were unavailable -- and the existing FAQ page was static, non-conversational, and quickly became outdated.
Purpose-built LLM chatbot for the study-abroad journey -- course discovery, shortlisting, document checklists, visa guidance, scholarship eligibility -- in natural multi-turn dialogue with live data.
🎨
Design, Content & Flow
  • Defined 40+ intent categories across the full student journey lifecycle
  • Wrote conversation flows, fallback scripts, and escalation path content
  • Designed mobile-first chat UI with proactive trigger logic and contextual cards
  • Built prompt library covering all intents with tone and format guidelines
🚀
Build, Deploy & GTM
  • LLM integration with live institutional API feeds for real-time accuracy
  • Embedded in consumer app and website with contextual proactive triggers
  • Weekly conversation quality audits to iterate on prompts and flows
  • Measured: CSAT, containment rate, escalation-to-counsellor ratio
24/7 student supportContainment rate improved1M+ users served
LLMMulti-turnIntent DetectionLive API
💬
04
GenAI Platform · Edvoy Production
GenAI Content Engine
Problem Statement
Edvoy needed high-quality, localised content across 750+ institutions and 90+ markets. Manual content creation could not scale -- the team was a bottleneck to both SEO performance and student information quality.
GenAI content platform that auto-creates, localises, and personalises educational content at scale -- university profiles, articles, student guides, and email campaigns with SEO baked in.
📝
Content Architecture & Design
  • Defined taxonomy across 12 content types and 90+ markets
  • Built prompt libraries per content type with brand voice guidelines
  • Designed human review workflow with quality gating stages before publish
  • SEO scoring layer integrated into generation pipeline
🚀
Build, Deploy & GTM
  • Integrated with CMS for one-click publish post AI generation and review
  • Tracked organic traffic lift comparing AI-generated vs human-authored content
  • Launched with content team onboarding and editorial quality guidelines
10x content outputConsistent brand quality at scaleSEO-optimised across 90+ markets
GenAIContent AutomationLocalisationSEOCMS
05
Personal AI Build · Full-stack Commerce
AI E-Commerce Platform with Order Management System
Problem Statement
Small and mid-size sellers lack access to a modern, AI-enhanced commerce stack that supports cross-border selling, intelligent product discovery, and end-to-end order management -- without the complexity and cost of enterprise platforms. Existing tools either over-engineer the tech or under-deliver on seller experience.
A complete e-commerce platform built from scratch with a fully integrated OMS -- AI-powered discovery, personalised recommendations, dynamic pricing, cross-border seller configuration, and an end-to-end order lifecycle. Built as part of consulting work across retail and e-commerce clients.
🎯
Planning & Architecture
  • Defined buyer-side and seller-side surfaces including cross-border seller config
  • Mapped OMS modules: order ingestion, status FSM, fulfilment routing, returns, refunds, analytics
  • Headless architecture: decoupled Next.js storefront plus OMS microservice
  • Seller config: multi-currency, cross-border shipping zones, tax rules, customs docs
🎨
Product Design -- Buyer, Seller & Ops
  • Buyer journey: home -- AI search -- PDP -- cart -- checkout -- confirmation -- tracking
  • Seller dashboard: catalogue management, cross-border pricing config, inventory alerts
  • Seller cross-border config: currency, shipping carrier per region, tax and customs automation
  • OMS operator UI: order queue, exception handling, bulk actions, SLA tracking board
Technical Build & Environments
  • Storefront: Next.js with SSR for SEO and Core Web Vitals
  • OMS backend: Node.js microservices, PostgreSQL, Redis for order state management
  • AI: collaborative filtering plus LLM-enhanced semantic search recommendation engine
  • Cross-border: multi-currency engine, shipping zone matrix, customs document generation
  • Docker plus Kubernetes, staging and production environments, full CI/CD pipeline
🚀
Deployment & GTM
  • Deployed on AWS: ECS plus RDS plus CloudFront CDN with multi-region redundancy
  • Onboarded first merchant as part of consulting engagement -- full seller setup, cross-border config, and catalogue migration
  • GTM: SEO-first content launch, Instagram and Pinterest organic, influencer seeding
  • KPIs: GMV, conversion rate, cart abandonment, OMS SLA adherence, return rate, cross-border order %
Full commerce stack built soloCross-border selling supportedLive with consulting clientEnd-to-end OMS with returns and analytics
Next.jsNode.jsPostgreSQLRedisAI RecsCross-borderOMSAWSK8s
🛒
06
Personal AI Build · Productivity Tool
Prompt Generator -- Cross-Industry AI Engine
Problem Statement
Most professionals adopting AI tools struggle to write effective prompts for their specific domain context. Generic prompt templates fail to account for industry nuances, role-specific language, and structured output requirements.
Structured prompt generation platform spanning 20+ industries -- marketing, legal, finance, HR, product, engineering, healthcare, real estate. Context-aware prompts with variable injection, tone calibration, and multi-tool export.
🎯
Planning, Content & Design
  • Mapped 20+ industries and 100+ use-case categories requiring structured prompt support
  • Defined prompt taxonomy: role, context, task, format, tone, output constraints
  • Designed variable injection system for personalised prompt instantiation
  • Content-wrote 500+ base templates across all industry verticals
🚀
Build, Deploy & GTM
  • Prompt composition engine with LLM-powered refinement suggestions
  • One-click export to ChatGPT, Claude, Gemini, and Copilot
  • Deployed as web app: free tier plus premium industry packs
  • GTM: Product Hunt launch, LinkedIn creator content, SEO prompt library pages
500+ templates live20+ industries coveredMulti-tool export
LLMPrompt EngineeringTemplate EngineReactSEO
07
Personal AI Build · Enterprise HR SaaS
AI-First HR Platform
Problem Statement
SMBs cannot afford enterprise HRMS software, yet their HR teams are buried in manual workflows -- leave approvals, onboarding checklists, performance reviews, and policy queries that could be fully automated or AI-assisted. No existing tool serves all three stakeholders (employee, manager, HR) in one coherent product.
Comprehensive HR SaaS with three product modules: Employee Self-Service, Organisation Management, and HR Operations. AI layer powers performance analysis, attrition risk scoring, and an HR policy Q&A chatbot.
🎯
Planning & Module Architecture
  • Three product surfaces: Employee app, Org admin console, HR ops dashboard
  • Mapped 25+ HR workflows: onboarding, leave, attendance, payroll sync, appraisals, exits
  • AI scope: performance trend analysis, attrition risk scoring, policy chatbot
  • RBAC architecture: employee / manager / HR / admin tiers
Build, Deploy & GTM
  • Node.js plus PostgreSQL with multi-tenant schema and row-level security
  • AI: RAG-powered HR policy Q&A plus attrition risk ML model
  • GDPR-compliant: field-level PII encryption, audit logs, data retention policies
  • Freemium: free up to 25 employees, paid tiers for larger headcount
  • Self-serve onboarding: company setup under 30 minutes
25+ HR workflows automated3 user surfaces in one productAI attrition scoringSelf-serve SMB SaaS
Node.jsPostgreSQLMulti-tenantRAGRBACAWSGDPR
🏢
04 / Toolkit

Deep craft.
Broad range.

01
🧠
AI & Machine Learning
RAG SystemsAgentic PipelinesLLM IntegrationGenAI PlatformsPrompt EngineeringVector DBsLangGraphSemantic Search
02
🏗
Product & Strategy
0 to 1 ProductEnterprise SaaSPLG StrategyConsumer AppsMulti-sided PlatformsERP ArchitectureRoadmappingPricing Strategy
03
💳
Fintech & Payments
Payments InfrastructureEdFintechCross-border FinanceLending ProductsBanking API IntegrationFX & Multi-currency
04
Engineering Collaboration
API DesignSystem ArchitectureMicroservicesCloud & DevOps (AWS)Docker / K8sCI/CDObservability
05
📊
Data & Analytics
Product AnalyticsFunnel OptimisationA/B ExperimentationSQL / Data QueryingDashboardingCohort Analysis
06
🌎
Leadership & Scale
Cross-functional LeadershipStakeholder ManagementGlobal Product StrategyTeam BuildingP&L OwnershipOKR / KPI Frameworks

Building
something
meaningful?

I work with teams and founders building products that genuinely matter -- across AI, fintech, edtech, enterprise SaaS, and beyond. If you have a complex problem worth solving, let's talk.

Start a conversation