Teams make product, brand, and marketing calls without hearing how customers really talk about their problems.
Build
A research pipeline that reads public conversations at scale, finds themes, and determines the sentiment of each theme.
Result
Auditable consumer insights traceable to source materials.
How it works
My judgment
Python
Claude reads
Deliverables
Out of the run
Key decisions
An auditable pipeline that reads what a market says when nobody is asking, then themes, sizes, and scores conversations for sentiment.
A human writes the brief: the communities, time window, and what counts as relevant. The pipeline runs from there in stages.
Four decisions:
Themes are discovered from the corpus, not supplied to it.A stratified discovery sample first, then full-corpus mapping against what came back.
Sentiment is scored per aspect named in an item.So one post that praises the price and attacks the side effects is counted as both.
Scoring is arithmetic, never the model's opinion.Confidence comes from sample size, source spread, temporal consistency, internal agreement, and recency; signal strength comes from prevalence, engagement, intensity, and conversation depth.
Every emotion score prints beside real quotes from the items scored that way.So the classification can be audited rather than trusted.
My background in the humanities taught me that how you frame a question decides the answer you get, that the loudest voices in a dataset are rarely the representative ones, and that a claim is worth only as much as the evidence you can put next to it. Every build I ship publishes its own limitations for the same reason.
The repo lists 14 known limitations, each with a severity and whether it is fixable.
Product teams can use this engine to find the language customers already use. Marketing teams can use it to create relevant content.
A regulated product can’t be shared socially, so growth has to come from 1:1 loops, content and the customer experience, with as little friction as possible.
Build
A growth system for the product: a 1:1 referral flow that captures an identifying contact, funnel instrumentation that catches breaks, claim resolution, and personalised content that solves for adherence and retention.
Result
Smoother referrals, better customer experience, fewer support tickets, with the infrastructure to retain customers for longer so they refer more.
How it works
Start and end
Friend's journey
Claim created
Out of the run
Key decisions
A UK GLP-1 provider wants to take its 30-day K factor from 0.0835 to 0.30.
I started with the target metric. Their own numbers say the referral does not happen at checkout: 1.21% share within 1000 seconds of paying, first referral at 26.8 days. The nature of the product could explain why: the starting dose is held for four weeks, and visible weight loss usually comes after that. A 30-day K is measuring patients before the medicine has done anything worth talking about, so it is worth interrogating whether a discount on the next order is enough incentive to push customers to refer before they may be ready.
Five builds:
Claim-first contact capture.The friend claims the discount with a phone number or email as step one, so the referral survives a device switch, a refused cookie banner and an abandoned signup. Review analysis showed lost claims are a live complaint today.
Funnel instrumentation: 9 events.3 for the sender, 6 for the friend, joined by 4 ids that hand off in sequence. Helps determine where the funnel breaks.
Referral resolution in Postgres.When a referred friend signs up, the system decides whether they are new or already in the database. Email and phone are unique, so it matches on the four remaining fields: first name, last name, address line, and postcode. 12 of the 16 combinations resolve on their own; 4 go to a human to check against the ID record. Claims go through cleanly, which cuts support tickets.
Content engine.The 13 structured data points collected at onboarding drive short, clinician-approved pieces delivered through an app.
Freemium app, 4 tabs.Dose reminders and refill prompts, a weekly content piece, weight and habit tracking, and a referral tab that sends a friend a GLP-1 primer along with the link. It gives a customer a reason to remember the product between deliveries.
Sales teams waste time on lists that are stale, wrong, or full of existing customers.
Build
An outbound system that spots buying signals in trustworthy public data, scores each account, finds the right contact, avoids spam and duplication, maintains a source of truth, and updates a CRM.
Result
Communication sent to the right candidate accounts, with no duplicate outreach, and every decision documented and the pipeline auditable.
How it works
Python and Supabase
Clay
n8n and Pipedrive
Out of the run
Key decisions
Built for Mercury, the fintech for startups, based on its own ICP definition: startups that just raised.
I went looking for the best signal of the funding event and settled on the SEC's Form D.
Mandated by law within 15 days of the first sale, so it is timely and comprehensive
Non-optional fields carry the company's legal name, location, amount raised and key personnel, ripe for enrichments
It is machine-readable through a free public API for automation
20 days of filings produced an average of 41 targets a day after cleaning.
Scoring decides resource allocation, and I am proudest of the scoring system out of 10:
Signal
Points
Why
Cash on hand
5
A proxy for cash on hand, plotted on a log curve taken from Mercury's inbound form, on the logic that more money on hand means a more valuable customer.
Raise completed
1
For companies that have most of their raise completed and are presumably closer to their business banking decision. It also spreads the scores out, and it penalises a company in a good industry that announced a large round but has raised little of it.
Industry
3
Mapped to the SEC's own values, seeded by assumed lifetime value for the demo and drawn from real customer data in production. The assumption is that Mercury's offering suits some segments better than others.
Filing history
1
Penalises companies that have raised many times before, on the assumption that their banking question was settled long ago.
Other judgement calls
A signer who filed for several companies collapses into one email about their largest raise, and the rep sees how many they signed for, so nobody gets spammed
Clay's stock waterfall got domains wrong, so Claygent runs with ranked corroboration candidates and returns its evidence, a confidence score, and the candidates it rejected
Every enriched row is deduped against existing customers, inbound, and every address the campaign has written to, seeded for the demo
n8n creates the organisation, person and deal in Pipedrive, then writes all three IDs back to Supabase, which stays the source of truth.
2,953 companies ingested, 830 scored. Every decision is documented, so a client can run this without me.
I work across the whole funnel: I find out what customers actually want, build the product loops that make them stay and refer, and build the outbound that brings in more of the right customers.
At a consumer fintech, I took acquisition from under 100 to 35,000+ paying customers in 18 months against funded competitors. About 40% came through referral, and activation rose 30%.
I build the systems myself, not just the strategy: research pipelines, databases, tracking, and outbound automations.
Areas of expertise
GTM ENGINEERING AND AUTOMATION
Systems ArchitectureProgrammatic ingestion pipelines, Event schema design, Attribution tracking
Infrastructure and OrchestrationPython, SQL, Supabase, Clay, n8n, Zapier, Apify, Claude Code
Data ModellingRelational schemas, Data integrity layers, Identity resolution and deduplication
BuildsNext.js, React Native, Rapid prototyping
PRODUCT LED GROWTH AND RETENTION
Funnel OptimisationOnboarding architecture, Activation loops, Experiment design and CRO
Strategic ExecutionGTM strategy, ICP discovery, JTBD research, Pricing and packaging
PerformanceCAC control, ROAS management, Paid search and social
Professional experience
May to Aug 2026Pune, India
Product and Growth Consultant
Animal Medical Services
Veterinary health platform: the commercial strategy and the first product build.
GTM Strategy: built the commercial model: competitive analysis, regulatory mapping, market sizing, and unit economics to define market entry criteria.
Voice AI: built a voice-first vet dashboard in English and Marathi: the vet speaks a clinic update, AI transcribes and structures it into a draft, and translates it into the other language for the vet to approve.
Full-Stack Prototyping: built and shipped a React Native app on a Supabase backend, solo with AI.
Nov 2025 to Apr 2026Remote
Growth and Product Lead
Consuma AI
Early-stage B2B SaaS consumer insights platform: pricing and productisation.
Growth Architecture: designed the productisation roadmap, pricing model, GTM framework, and growth metrics system.
Product-Market Fit Strategy: designed JTBD-based user presets and output standardisation workflows to capture product-market fit signals.
Operational Leadership: directed a small team and external contractors across product, growth, brand, PR, and operations.
Jun 2024 to Nov 2025Remote
Marketing Manager
ScopeX Fintech
Mobile-first consumer fintech serving EU markets.
Hyper-Scale Acquisition: scaled the funnel from under 100 to 35,000+ paying customers within 18 months against funded and legacy incumbents; referrals drove roughly 40% of new customers.
Funnel Optimisation: raised activation 30%+ through onboarding and UI experiments, and grew revenue 133%+ through lifecycle marketing in HubSpot.
Capital Efficiency: tripled ROAS with sharper targeting and narrative-driven landing pages, and held CAC down through lean paid acquisition, partnerships, and CRO.
Data Infrastructure: built the event schema and reporting that showed which channel drove signups and retained revenue.
Cross-Functional Operations: served as pseudo-GM: fixed payment flows and compliance gaps, managed a small team of marketers, and built support ops that held CSAT above 85 during aggressive growth.
May 2017 to Sep 2026Pune and Remote
Founder, Clinical Nutrition and Lifestyle Design
BioRhyme Health
Multi-disciplinary metabolic health practice, founded with a consultant psychiatrist and a general physician.
Practice Growth: grew the practice to 300+ long-term clients, effectively all acquired peer-to-peer or by GP referral.
Retention: improved retention with habit and engagement loops, and daily progress tracking.
Behavioural Research: 9 years of direct client interviewing on why people start treatment, why they stop, and what makes them refer.
Dec 2015 to Mar 2019Pune, India
Business Development and Marketing
Edvance Pre-schools
Regional B2B sales, admissions ops, and enrolment marketing events across five preschools.
Jun 2012 to Sep 2015
Artist Management and Events
Self-employed
Artist representation, brand sponsorships, corporate and hospitality events, and national tour logistics for touring musical acts.
Education
MA, Gender, Culture and Development Studies, Savitribai Phule Pune University, 2019 to 2021
BA, Media and Communications, Symbiosis Centre for Media and Communication, 2008 to 2012
Technical Programs: Full-stack Developer and AI Growth Engineering (Scrimba, in progress) · GTM Engineering (StackOptimise) · Digital Psychology and Growth (CXL Institute) · Google AI Essentials and UX Design Professional Certificate
Languages: English (native) · Hindi (native) · Marathi (professional working) · Mandarin Chinese (limited working)
Header painting: Hilma af Klint, Group X, No. 1, Altarpiece, 1915, Hilma af Klint Foundation, Stockholm.
Your website, your ads and your sales pitch each describe your product a little differently, and none of it sounds like your customers. I find out who buys from you, why, and what they compared you with. Then I turn it into one clear story your whole team can use.
You sell to
Stage 1
Learn how you sell. First, I learn how your business works: what you sell, who buys it, and how you sell it. I read your customer data and listen to your sales calls. I write the questions for your customer interviews and record a training, so your team can run them. And I start reading what your customers, and your competitors' customers, say in public: reviews on G2 and Capterra, forums, LinkedIn.
Stage 1
Learn how you sell. First, I learn how your business works: what you sell, who buys it, and how you sell it. I read your customer data and your support conversations. I write the questions for your customer interviews and record a training, so your team can run them. And I start reading what your customers, and your competitors' customers, say in public: app store reviews, Reddit, forums, social media.
Stage 2
Hear it from your customers. Next, your team runs the interviews, as many as it takes. Meanwhile, I study your competitors: what they promise, where they're strong, and where they fall short.
Stage 3
Find why they choose you. Now I put it all together: who buys from you and why, in their own words, and what they compared you with. From that, I write your positioning: who the product is for, what it beats, and why.
Stage 4
Make it yours. Finally, I turn the positioning into your story and your messages: one story for the company, and the right message for each kind of customer.
What you keep
Your positioning, your company story, and the messages for each kind of customer, with who your best customers are and why they buy. It's all in one document, with the evidence behind every line.
You can also add
Your sales pitch, a one-pager, and a battlecard for each main competitor, showing how to win against them; two versions of your homepage, layout and copy, ready to test against each other; case studies written from your customer interviews.
The project gives you your positioning, your story and your messages. But words in a document don't win customers. They have to show up on your website, in your ads and in your campaigns. If I stay on, I put them there, one change every week.
Where the project stops
You have your positioning, your story and your messages in one document. Putting them on your website, in your ads and into a brand falls to your team, on top of everything else.
If I stay on
I grow one number: how many of your website visitors sign up, book a demo or start a trial. We agree a target for it from your data, and I ship a change for it every week.
Every week
✓We pick the change. Once a week we meet, look at last week's result, and pick the next change from the list.
✓It gets made. Your designer makes it from my direction, and your engineers put it live. If it's words, I write them myself.
✓It goes live, measured. I check the tracking works the day it ships.
✓I read what happened. Some tests need a few weeks of visitors before they show a winner. The result tells us what to change next. You get a short report before each meeting.
What a change can be
A page on your website, rewritten and tested; new messages and visuals for your ads, tested for whoever runs them; a brand campaign or a launch, made with your designer.
Your brand, built alongside
A brand takes longer than a week, so it's built alongside, in parts: your voice first, then your look, then the guidelines your team follows. Each part shows up in the weekly changes as soon as it's ready. A brand pays off slowly, so I don't judge it by sign-ups alone.
Every quarter
I send you a short written review: where the number stands against its target, what changed, and what comes next. It also shows how your brand is doing: how many people search for you by name, how many come straight to your site, how many subscribe, and what people say about you in reviews.
You can also add
Content every month in your brand's voice, for each channel you choose: blog posts, social posts, and posts for your founder; your whole website rebuilt on your positioning; your sales pitch, a one-pager, and a battlecard for each main competitor, showing how to win against them; case studies written from your customer interviews.
Find out what's holding your growth back, and fix it
You've raised, and growth isn't following because something between sign-up and renewal is broken. Some of your customers already do exactly what you want all of them to do. I study them, and get more of your customers to do the same.
You sell to
Stage 1
Find the real problem. First, I learn how your business works: how you make money, how you sell, and how your customers find you, pay you and talk to you. I also read what your customers, and your competitors' customers, say in public: reviews on G2 and Capterra, forums, LinkedIn. Then I bring you four or more problems worth solving, each with its evidence. You pick one, or we let the data decide.
Stage 1
Find the real problem. First, I learn how your business works: how you make money, how you sell, and how your customers find you, pay you and talk to you. I also read what your customers, and your competitors' customers, say in public: app store reviews, Reddit, forums, social media. Then I bring you four or more problems worth solving, each with its evidence. You pick one, or we let the data decide.
Stage 2
Study the customers who got it right. Next, I look at the customers who already do what you want. Where did they come from? What did they do when they first signed up? Who did they bring with them? What sets them apart? Then I compare them with the customer cohorts who don't take the steps you want.
Stage 3
Choose the fix. Now I list every fix for that problem, ranked by how much it matters and how far it could move the number. You choose one that can go live right away, and I build the mock-up.
Stage 4
Ship it. Finally, the fix goes live, measured from day one. If it touches production code, your engineers build it from my mock-up. If it doesn't, I build it myself.
What you keep
One fix live and measured, and one document with everything else: how your business works, the problems we found, what your best customers do differently, every fix ranked, and what to do next.
You can also add
Your pricing, packaging and communication reviewed against publicly available competitor data; content that tells your customers about the change: marketing material, blog posts, LinkedIn posts, and posts for your founder.
You can also add
Your pricing, packaging and communication reviewed against publicly available competitor data; content that tells your customers about the change: social posts, your app store listing, and the emails and in-app messages that announce it.
The project finds what's holding your growth back and ships the first fix. But it finds more than one fix, and your team is busy building your product. So the other fixes wait. If I stay on, I keep shipping them, one every week.
You sell to
Where the project stops
You have one fix live and a ranked list of the rest. Building them falls to your team, on top of everything else they're doing.
If I stay on
I pick one number to grow, like how many trials become paying customers, or how many customers renew. We agree a target for it from your data, and I ship a fix for it every week.
If I stay on
I pick one number to grow, like how many sign-ups start paying, how many customers come back, or how many new customers come from friends. We agree a target for it from your data, and I ship a fix for it every week.
Every week
✓We pick the fix. Once a week we meet, look at last week's result, and pick the next fix from the list.
✓It gets built. If it touches production code, your engineers build it from my mock-up. If it doesn't, I build it myself.
✓It goes live, measured. I check the tracking works the day it ships.
✓I read what happened. The result tells us what to fix next. You get a short report before each meeting.
What a fix can be
The emails your customers get after they sign up, in-app messages, a cancel flow that gives customers a reason to stay, recovering failed payments, your sign-up and pricing pages, or a product change I mock up for your engineers. Bigger changes get built alongside, while the weekly fixes keep going.
What a fix can be
The emails, push notifications and in-app messages your customers get after they sign up; a cancel flow that gives customers a reason to stay; recovering failed payments; invite-a-friend and sharing; your app store listing and the first screens a new customer sees; your sign-up and pricing pages; or a product change I mock up for your engineers. Bigger changes get built alongside, while the weekly fixes keep going.
One number at a time
When the number reaches its target, or stops moving, we move to the next one. Every quarter I send you a short written review: where the number stands against its target, what shipped, and what comes next.
You can also add
A pricing or packaging change, researched, tested and shipped; a launch for a new feature, plan or market; the job description for your first growth hire; content every month about what shipped, for each channel you choose.
You can also add
A pricing, plans or paywall change, researched, tested and shipped; a launch for a new feature, plan or market; the job description for your first growth hire; content every month about what shipped, for each channel you choose; your ad messages and visuals tested every month, for whoever runs your ads; a creator programme: the brief, how to choose creators and the terms; your community planned and launched.
You've raised, and your product is ready to sell. But finding buyers still means someone searching by hand, one company at a time. I build you a system that finds the right companies the moment they're ready to buy, and puts each one in front of your team.
Stage 1
Choose who to go after. First, I learn how your business works: what you sell, who buys it, and why. I study the deals you won and the ones you lost, and I read what your market says in public: reviews, forums, your competitors. Then we choose one group to go after: one kind of buyer, or one kind of partner who can bring you buyers. I also set up the inboxes you'll send from and start warming them up, so your emails land in the inbox, not in spam.
Stage 2
Find the moment they're ready. Next, I turn that group into a description a database can search. Then I find the signs that show a company is ready to buy, and where each sign is published.
Stage 3
Build the system. Now I build it on your own accounts. Every day, it finds the companies in your group, leaves out the ones you can't serve, scores the rest, and finds the right person to write to. It never contacts your customers, anyone who already came to you, or anyone you've already written to. Every company lands in your CRM, and I write the first email and the follow-ups.
Stage 4
Hand it over. Finally, I test it on a real list, with nothing sent. When someone replies, their deal moves forward in your CRM on its own, and your forecast updates as deals close. Whoever answers replies and takes the calls gets a recorded training. Once the inboxes are warm, the system starts sending on its own, and your team handles the replies.
What you keep
A working system that finds your buyers every day, on your own accounts, and one document that explains every part of it: who you're going after and why, the signs that tell you they're ready, how each company is scored, and the emails. It's yours to run, with or without me.
You can also add
A second group, built the same way; LinkedIn messages alongside the emails; live training on answering replies and taking calls, instead of the recording.
The project builds a system that finds your buyers every day. But a system like this is never finished: markets move, messages wear out, and new kinds of buyers show up. If I stay on, I keep making it better, one change every week.
Where the project stops
You have a working system on your own accounts. Finding new buyers, testing new emails and tuning the scoring falls to your team, on top of selling.
If I stay on
I grow one number: how many real sales conversations you have each month with the right buyers or partners. We agree in writing what counts as a real conversation, and a target for it from your data. The system sends on its own; your team answers the replies and takes the calls.
Every week
✓We pick the change. Once a week we meet, look at last week's result, and pick the next change from the list.
✓I build it. It goes into the system on your own accounts, so it stays yours.
✓It goes live, measured. I check it works the day it ships.
✓I read what happened. The results tell us what to change next. You get a short report before each meeting.
What a change can be
A new sign that a company is ready to buy; scoring that learns from the deals you actually closed; new emails and follow-ups, tested against each other; a new kind of buyer or partner; scripts for whoever answers replies and takes calls, and coaching for them. Bigger changes run alongside, while something small ships every week.
Every quarter
I send you a short written review: where the number stands against its target, what changed, and what comes next.
You can also add
A new country or market, with its own buyers, signs and emails; LinkedIn messages alongside the emails; sales material built from your positioning: a one-pager, a deck and case studies; the job description for your first salesperson.
