// data analyst · analytics engineer

Data in.
Decisions out.

I am a Data Analst and Anaytics Engineer with Bachelor's in Computer Science. History of uncovering ~$10M in revenue leakage, reducing costs by ~$1000/month, increasing marketing ROI by 20%, and signups boost by 25%. Cross-collaborated across marketing, sales, product, and growth, building tailored solutions for thier success. Managed pipelines, turned raw data into insights and presented them to leadership including CEO, COO, CTO that moved the needle and helped businesses grow!

Explore my work Let’s talk
  • 5+ years in data
  • Worked with Startups, Scaleups, and Enterprises
  • Complete Time Zone Overlap for US, Europe, UK, UAE
impact_report.sql 
SELECT metric, value
FROM career_impact
WHERE analyst = 'Ujjwal Jha'
ORDER BY business_value DESC;
revenue_analyzed$100M+
revenue_leakage_surfaced~$10M
query_runtime_cut96.5%
pipeline_uptime99.99%
4 rows returned · 14 ms

$100M+

Revenue analyzed end-to-end

leadership revenue-protection dashboard

~$10M

Revenue leakage surfaced

broken down by dimensions that helped solve it

96.5%

Query runtime cut

6m 40s → 14s using CTEs, indexes, joins

99.99%

Production pipeline uptime

7+ sources flowing into PostgreSQL

// how I add value

Tools are table stakes.

TECHNICAL: SQL, Python (Pandas, statsmodels, SciPy), Statistical Analysis, Hypothesis Testing, A/B Testing, ETL, ELT, Data Warehousing, AI Automation.

TOOLS: Snowflake, dbt, AWS S3, Supabase, Sigma, Power BI, Fivetran, Git, Excel, HubSpot, GitHub, Jupyter Notebooks, Claude Code.

But you don’t hire an analyst for the stack. You hire one for what happens to the business next.

fn find_the_money()

Find the money

Revenue leakage, commission accuracy, payout forecasting. I dig until the numbers explain themselves, then hand leadership a decision, not a data dump.

~$10M leakage surfaced · $7M commissions at 100% accuracy

fn make_data_dependable()

Make data dependable

Pipelines, modeling, and warehousing that nobody has to babysit. Engineered for reliability first and cost a very close second.

99.9% uptime, Fivetran spend ~$1000/M → $0

fn ship_decisions()

Ship decisions, not just charts

Executive views the C-suite actually opens every morning. Designed around the questions leadership is really asking.

$33M+ tracked for COO/CTO · 3 dashboards in the first 4 days of joining

// where I’ve done it

Proof, in production.

Analytics Engineer

US-Based Fitness Company

Nov 2025 – Jan 2026

  • Identified ~$10M revenue leakage by building a revenue analytics solution, analyzing $100M+ in revenue and categorizing leakage by refund tiers, customer segments, and onboarding patterns to support revenue protection strategies.
  • Measured $33M+ in yearly performance across CPA, Gross, Net, and New Revenue, solving pipeline and data quality issues across HubSpot, payment processors, Google Ads, and Meta Ads, creating an Operations Dashboard in Sigma for the CTO, COO, and CEO.
  • Recorded ~$2M in sales and 300+ renewals during Black Friday sales by designing 3 real-time tracking solutions using Supabase and Sigma, showing sales performance via dashboards to Sales, Marketing, and C-level executives built within 4 days of joining.
  • Optimized SQL queries by 96.5% (6m 40s to 14s) used by business-critical data solutions, replacing inefficient statements, reducing complex joins, designing indexes and partitions, and improving all the data solutions for executive and operational users.
  • Designed a sales commission system tracking $7M+ in monthly sales and commissions for closers by creating a self-serve, tier- based solution with 100% accuracy, replacing an error-prone solution to calculate commissions and sales for closers and Finance.
  • Reduced Fivetran costs by ~$1000/month by redesigning transformation strategies, eliminating inefficient model runs, and optimizing pipeline execution, bringing transformation costs to zero.
  • Ensured 99.99% uptime across production ETL pipelines integrating HubSpot, Google Ads, Facebook Ads, Google Analytics, Calendly, JotForm, and APIs into Supabase using Fivetran and dbt
  • Identified $4M+ in amount owed to fitness coaches through 2031 by developing a data solution used by CTO and COO, working across messy and fragmented data, building client lifecycle logic and coach assignment in SQL, allowing owners to save taxes.

SigmaSQLPostgreSQLFivetranHubSpotSupabase

Data Analyst Consultant

Early Stage Startups

Mar 2021 – Present

  • Increased product sign-ups by 25% through funnel analysis using SQL and web analytics, identifying OTP verification as the primary conversion drop-off, and partnering with the mobile team to resolve a template bug.
  • Elevated ARPU by 15% by identifying a low-performing marketing platform with low LTV:CAC, retention, and high churn, rebalancing budget on a high-performing alternative, driving improved customer acquisition and revenue growth.
  • Improved data processing efficiency by 40% by building ETL workflows using Snowflake, Python, and SQL on 20M+ row datasets, architecting a raw-to-stage-core layered pipeline from AWS S3, using storage integrations and Streams for incremental loads
  • Boosted marketing ROI by 20% by building a marketing analytics solution from scratch, covering ETL, data modelling, and KPI- driven dashboard (ROI, ROAS, CAC), guiding budget allocation towards high-performing regions and channels.
  • Eliminated 40 hours/month of manual, error-prone Excel tracking by building SQL and Power BI solutions across 5 teams (sales, revenue, marketing, product, growth) via dynamic SQL views, auto-updating with zero manual intervention and 100% accuracy

SQLPythonSnowflakePower BIA/B Testing

// the toolkit

The stack behind the outcomes.

TABLE skills.analytics_experimentation

SQL (Advanced)Python · PandasstatsmodelsStatisticsA/B TestingHypothesis TestingFunnel AnalysisCohort AnalysisRetention & ChurnKPI Design

TABLE skills.data_engineering

ETL / ELTData PipelinesData ModelingData WarehousingAPIsFivetran

TABLE skills.warehouses_databases

SnowflakeBigQueryRedshiftPostgreSQLMySQLSupabase

TABLE skills.bi_visualization

SigmaPower BI · DAXPower QueryLooker StudioExcel

// selected work

Projects with a point.

Each one built end-to-end: raw data → pipeline → model → a dashboard someone can act on.

All projects →

// next step

Got data? Let’s make it decisive.

I partner with teams that want clarity — fully remote, async-friendly. If your dashboards raise more questions than they answer, we should talk.

Email me LinkedIn

· Working across US/EU and many other time zones