Open to APM and PM roles

Shiwal Singh

Associate Product Manager · B2B SaaS · AI and integrations

I turn support signal into shipped product. Four and a half years inside B2B SaaS support taught me to spot the problem that keeps coming back, write the case for fixing it, and stay on it until the fix holds.

Portrait of Shiwal Singh
Based inLucknow, India
6wks
JetSend from discovery to GA at Maropost
30%
Fewer inbound tickets after the HighLevel AI agent shipped
40to 25
Minutes of average handle time with an internal AI tool
2
Enterprise clients cited JetSend at renewal
Strengths: discovery from ticket data, PRDs, AI product work, API and integration depth Also: MBA, McKinsey Forward, 12 PM certifications
Selected work

Things I shipped, built and took apart

Each one shows the problem, my role and what changed. The three product launches have full case studies below.

Live MVPFounder and PMAI product

Loopback: a weekly product digest from support tickets

PMs at small B2B SaaS companies decide roadmaps without reading the support tickets that hold the best signal. Enterprise voice of customer tools cost tens of thousands a year. Loopback reads the tickets and sends three points every Monday: top pain point, top feature request and one churn signal.

Status
Live since Aug 17, 2026, pre pilot
North Star
80% weekly digest delivery rate
Pricing
Four tiers, Free to $199 a month
Market call
Corrected TAM from $4.2B to $16M to $31M

Built solo with AI coding assistants. I own the product decisions, specs and QA, and the assistants write the code. The PRD labels every claim as a hypothesis or validated.

Input
Support tickets via Excel or CSV import
AI synthesis
Structured output, with every quote required to trace to a real ticket
Delivery
Monday digest in Slack and email
01Top pain point 02Top feature request 03One churn signal
Case studies

How I work, shown on real launches

Same structure every time: context, problem, what I did, outcome. Open any one.

JetSend: from a ticket pattern to GA in under 6 weeks Maropost · Client Support Specialist · Dec 2023 to Sep 2024

Context

Maropost is a marketing automation and ecommerce platform. As a support specialist I handled enterprise client tickets and saw the same transactional messaging complaints come back from different accounts.

Problem

Failed sends, manual workarounds and no visibility when something broke. Each ticket got closed, but the gap stayed open and was eroding client trust.

What I did

  • Analyzed 80+ support tickets to confirm the pattern and size it.
  • Ran customer interviews alongside the PM and framed the need with Jobs To Be Done.
  • Used a MECE trade off analysis to decide what belonged in scope.
  • Wrote the requirements brief with user stories and acceptance criteria.
  • Coordinated product, engineering and QA through to GA, then handed it to the product team after launch.

Outcome

< 6 wksto GA
80+tickets analyzed
2enterprise clients cited it at renewal
AI agent for white label clients HighLevel · Customer Support Representative · Apr 2025 to Feb 2026

Context

HighLevel is a white label CRM and marketing automation platform. The accounts I supported were North American SMBs reached through white label partners.

Problem

Clients kept asking the same repeat questions and needed a faster first response than the queue could give.

What I did

  • Configured and validated AI powered chatbot workflows, including conversational flows and webhook triggered logic.
  • Validated the flows before they reached clients.
  • Shipped in 6 weeks and measured the effect on ticket volume and CSAT.

Outcome

30%drop in inbound ticket volume
20%CSAT increase
6 wksto ship

Results as communicated by the clients.

Internal RAG assistant for the support team HighLevel · Customer Support Representative · Apr 2025 to Feb 2026

Context

The support team handled a large product surface, including HIPAA and GDPR compliant accounts.

Problem

The team was losing hours to manual knowledge lookup for answers that were already documented.

What I did

  • Defined requirements and wrote the functional documentation for a RAG based chatbot.
  • Coordinated the build and the rollout to the team.
  • Tracked adoption, accuracy and handle time after launch.

Outcome

40 to 25minutes average handle time
+5%CSAT
Loopback: decisions behind a solo built product Founder and PM · Feb 2026 to now · PRD v3, public

Context

B2B SaaS teams of 5 to 50 people have support tickets full of roadmap signal and no affordable way to read it. Loopback is my attempt at that gap, from PRD to a live product.

Problem

Reading 500 tickets by hand takes about 8 hours. AI synthesis now does comparable work in about a minute for a few cents, so a self serve product at SMB prices became viable.

Decisions that mattered

  • Three point digest, not a dashboard. The output is a decision, so it stays opinionated and needs no login.
  • Support lead sets it up, PM reads it. Setup is designed to take under 20 minutes with no taxonomy work.
  • Excel and CSV import instead of live Freshdesk sync. I had no real API access, so I cut scope and kept the core loop testable.
  • Corrected my own market sizing. The first TAM of $4.2B did not hold. I rebuilt it to $16M to $31M, labeled it a founder estimate and left the correction visible in the PRD.
  • Scored 12 features with MoSCoW and RICE, then cut redundant Kano labels that changed no decision.

What I built

Landing page, five screen onboarding, login, dashboard with digest history, CSV upload, role based access with multi tenant data isolation, four tier pricing with backend plan enforcement, and a weekly digest on a schedule. Built with AI coding assistants on Next.js, Supabase and a Gemini API synthesis step, deployed on Vercel.

Experience

Four roles, one habit: find why it keeps happening

Customer Support Representative
Apr 2025 to Feb 2026
HighLevel · Remote, US market
  • Handle time down 37%, from 40 to 25 minutes, with a 5% CSAT lift, by defining requirements for an internal RAG based AI agent and tracking metrics after launch.
  • 30% fewer inbound tickets and 20% higher CSAT across SMB white label accounts, by configuring and validating AI chatbot workflows. Shipped in 6 weeks.
  • Resolved technical escalations on HIPAA and GDPR compliant accounts, and tracked NPS to surface recurring product friction.
Technical Support Engineer
Oct 2024 to Apr 2025
Vartopia
  • Integration troubleshooting guide adopted by 3 teammates in the first month, covering REST API, webhook and CRM connector failure patterns.
  • Sprint backlog input accepted from a support role, proposing automation of form value passthrough between CRM systems.
  • Reproduced bugs, isolated root cause with logs and debugging tools, and worked with engineering through triage and resolution.
Client Support Specialist
Dec 2023 to Sep 2024
Maropost
  • JetSend shipped in under 6 weeks, cited by 2 enterprise clients at renewal. Discovery covered 80+ tickets, Jobs To Be Done and MECE scoping.
  • Competitive analysis fed roadmap discussions, documenting Maropost against key alternatives.
  • Resolved sync and data mapping issues between connected ecommerce systems.
Analyst
Aug 2021 to Dec 2023
eClerx
  • Consistent data accuracy across large scale ecommerce operations for global retail clients, through structured QA and documentation habits.
  • Built pattern recognition and escalation discipline at high ticket volumes, treating every ticket as a signal.
MBA, International Business Management
Amity University, Noida
Jan 2024 to Apr 2026
Core competencies
Product discoveryPRDs and user storiesAcceptance criteria Roadmap planningBacklog prioritizationStakeholder management APIs and webhooksAgile and ScrumRoot cause analysis Product analyticsCustomer interviews
Tools I use
FigmaAhaCursorClaude LovableNotionHotjarSQL (working)
Proof

12 certifications, each one verifiable where a link exists

Each was earned to close a specific gap. Tap a certificate below to enlarge it.

McKinsey Forward Program
McKinsey.org · Jun 2026
Product Management with AI
PW Skills, NSDC, PwC · Jun 2026
Verify ↗
Technical Product Management
PMI, LinkedIn Learning · Jun 2026
Verify ↗
Building a Product Roadmap
PMI, LinkedIn Learning · Jun 2026
Verify ↗
Building a Product Strategy
PMI, LinkedIn Learning · Jun 2026
Verify ↗
Product Prioritization (PPC)
Product School · Apr 2026
Agile Foundations
PMI, LinkedIn Learning · Jun 2026
Verify ↗
Scrum Fundamentals Certified (SFC)
VMEdu, Vabro.ai · Nov 2024
Generative AI for Product Managers
LinkedIn Learning · Jun 2026
Verify ↗
Data Driven Product Management
NASBA, LinkedIn Learning · Jun 2026
Verify ↗
Building AI Products: Prototyping Essentials
LinkedIn Learning · Jun 2026
Verify ↗
Introduction to AI
LinkedIn Learning
Play

A product ships when it does not crash into itself

Steer, grow the backlog, do not double back. Click the board, then use arrow keys or WASD. On a phone, swipe or use the pad.

SCORE 0BEST 0
Click the board to start
Recommendations

People who worked with me

From colleagues, a manager and a senior product leader. All public and verifiable on LinkedIn.

"He took it upon himself to talk to stakeholders, understand what clients actually needed from transactional messaging, and turned that into structured requirements the product team could use. That kind of initiative is rare in a support role. Two enterprise clients actually brought up the JetSend improvements during their renewal conversations. He'd be a strong addition to any product team."

Shubham SharmaTechnical Support Development Liaison, Maropost. Senior to Shiwal.

"Shiwal was different. He'd notice patterns in what was coming in, flag recurring issues before they became a bigger problem, and actually think about how to reduce the noise. He'd document findings properly, build out response templates, and put together resolution guides the whole team ended up using. Our average handling time came down."

Shubham KumarBusiness Analyst, eClerx. Direct manager.

"Most people in support are focused on closing the ticket in front of them. Shiwal was always thinking one step ahead. He didn't just flag things and disappear. He'd follow through, make sure engineering had what they needed, check back on the fix, then go back to the customer himself. He's switching to product management now and it genuinely doesn't surprise me."

Naveen PabrejaCustomer Success Manager, Maropost. Senior to Shiwal.
From a VP of Product and Engineering

"I worked closely with Shiwal at Vartopia, where he was a Technical Support Engineer. Shiwal was dependable and thoughtful, and what stood out was his curiosity about the product. He went beyond resolving support issues, he looked for root causes, identified recurring problems, and brought useful feedback to product and engineering. I'm confident he has the mindset and work ethic to grow into a strong product manager. He asks the right questions, thinks beyond the immediate problem, and follows through."

Atul KhindriaVP of Product and Engineering, Vartopia. Senior to Shiwal, not a direct manager.View on LinkedIn ↗

Read all recommendations on LinkedIn ↗

Contact

Hiring a PM who starts from the customer signal? Let's talk.