🟢 Open to work
I build financial models — and the tools that make them faster.
Models built from scratch. Numbers traced to the source. Recommendations I stood behind.
Picked the target. Built the model. Defended the call. Proposed Rubrik, Inc. (NYSE: RBRK) to my graduate team, then led the DCF and comparable-company valuation — revenue build, margins, discount rate, terminal value — in S&P Capital IQ, Bloomberg Terminal, and Morningstar. We pitched the buy recom
Selected to represent the Endowment model and we did Stanford University. Took apart a $40.8B endowment's asset allocation and multi-year risk-adjusted performance, benchmarked it against the Yale model, and presented the findings to Crystal Capital.
A live $9.74M, ten-position derivatives book. Priced it, stress-tested it, and hedged it — cutting cumulative hedging P/L variance by 92%.
Two issuers that could not be more different: a $29.4B state-owned company and a $75B private one. Compared every route to raising capital and recommended a structure for each.
A $1,000,000 mandate with a real constraint set. Built the allocation, wrote the policy, and monitored it against defined risk targets rather than just reporting returns.
Stopped pulling data by hand. Built a Python pipeline to fetch it automatically, then used R to work out which factors were actually driving returns and which were noise.
Two public companies, multi-year SEC filings, no shortcuts. Built the full analysis from primary documents and traced the performance gap to the specific line items causing it.
Selected for a live-client operations case. Analyzed patient-flow and capacity constraints at a major health system and quantified what fixing them was worth.
I don't just use AI — I build with it and I check its work. Deployed an agent that runs unsupervised against a live account, and I'm building an app that scores how deeply AI is integrated into a workflow. In financial analysis, AI speeds up my first pass; verification decides what survives.
I build financial models — and the tools that make them faster.
I'm Sarah Noor. I found finance by running a business, not by studying it first — which means I learned what a receivable is by chasing one. Two years working on a $2–3M P&L, then an MS in Applied Finance at Pepperdine. Since: a valuation defended in front of real VC investors, a $9.74M derivatives book hedged, a $40.8B endowment picked apart for Stanford, and one AI agent that handles my inbox. I check everything twice. It's a personality trait, not a policy.
Experience
Optometrist / Clinical Operations Associate
Khanna Vision Institute · Jun 2023 Sep 2025
Research Assistant, National Public Health Study (Phase II)
AIIMS / Indian Council of Medical Research (ICMR) · Dec 2018 Jun 2019
Education
Pepperdine University, Graziadio Business School
Master of Science in Applied Finance · Class of 2026
UCLA Extension
Coursework: Applied Managerial Finance (MGMT X 130A), Ethics in Finance (MGMT X 441)
All India Institute of Medical Sciences (AIIMS)
Bachelor of Optometry · Class of 2018
Skills
Picked the target. Built the model. Defended the call. Proposed Rubrik, Inc. (NYSE: RBRK) to my graduate team, then led the DCF and comparable-company valuation — revenue build, margins, discount rate, terminal value — in S&P Capital IQ, Bloomberg Terminal, and Morningstar. We pitched the buy recom
Picked the target. Built the model. Defended the call. Proposed Rubrik, Inc. (NYSE: RBRK) to my graduate team, then led the DCF and comparable-company valuation — revenue build, margins, discount rate, terminal value — in S&P Capital IQ, Bloomberg Terminal, and Morningstar. We pitched the buy recommendation live to Bryant Stibel's investment professionals and defended it under questioning.
Sourced the idea myself — screened the sector, proposed Rubrik, Inc. (NYSE: RBRK), and the team ran with it. Built the DCF and comparable-company valuation end to end: revenue build, margin assumptions, discount rate, terminal value. Wrote the investment thesis covering market opportunity, trading behavior, and macro risk.
Delivered a complete valuation and written buy recommendation to a live audience of practicing VC/PE investors — and the thesis held when they pushed on the assumptions. The exercise taught me the difference between a model you submit and a model you have to justify out loud, which is the standard I've built everything since to.
Selected to represent the Endowment model and we did Stanford University. Took apart a $40.8B endowment's asset allocation and multi-year risk-adjusted performance, benchmarked it against the Yale model, and presented the findings to Crystal Capital.
Selected to represent the Endowment model and we did Stanford University. Took apart a $40.8B endowment's asset allocation and multi-year risk-adjusted performance, benchmarked it against the Yale model, and presented the findings to Crystal Capital.
Analyzed allocation across public and private markets, including alternatives exposure, and measured multi-year risk-adjusted performance against the Yale endowment framework. Authored a 10-page written report and built the presentation.
Presented findings and recommendations directly to Crystal Capital Partners' investment professionals and faculty. The work is the closest thing in my portfolio to an institutional allocation review — a real framework, a real benchmark, and an audience that allocates capital for a living.
A live $9.74M, ten-position derivatives book. Priced it, stress-tested it, and hedged it — cutting cumulative hedging P/L variance by 92%.
A live $9.74M, ten-position derivatives book. Priced it, stress-tested it, and hedged it — cutting cumulative hedging P/L variance by 92%.
Bootstrapped SOFR par swap rates to construct the discount curve and built a three-period binomial lattice off a 34.04% annualized volatility estimate to price optionality. Quantified the book's exposures, stress-tested across market scenarios, then extended a delta-hedging program to delta-gamma hedging.
Cumulative hedging P/L variance fell 92% across tested scenarios. It's the most technically demanding work in this portfolio and the piece I'd point to for anyone asking whether I can handle complex securities.
Two issuers that could not be more different: a $29.4B state-owned company and a $75B private one. Compared every route to raising capital and recommended a structure for each.
Two issuers that could not be more different: a $29.4B state-owned company and a $75B private one. Compared every route to raising capital and recommended a structure for each.
Researched the end-to-end IPO process — S-1 filing requirements, roadshow and bookbuilding, pricing, and greenshoe stabilization — against direct listing, auction, and SPAC structures, weighing the regulatory and market mechanics behind each.
Delivered a tailored financing recommendation for both issuers, grounded in the deal mechanics rather than a generic preference for going public. It's the project that built my working knowledge of how capital markets transactions actually get executed.
A $1,000,000 mandate with a real constraint set. Built the allocation, wrote the policy, and monitored it against defined risk targets rather than just reporting returns.
A $1,000,000 mandate with a real constraint set. Built the allocation, wrote the policy, and monitored it against defined risk targets rather than just reporting returns.
Constructed a diversified, investment-policy-compliant multi-asset portfolio, then tracked performance against defined targets using beta, Sharpe and Sortino ratios, and maximum drawdown. Produced a documented allocation and rebalancing policy governing ongoing compliance.
A complete, documented mandate — allocation, policy, and recurring risk-adjusted performance reporting — rather than a static asset mix. It's the project that taught me portfolio management is mostly governance, not stock picking.
Stopped pulling data by hand. Built a Python pipeline to fetch it automatically, then used R to work out which factors were actually driving returns and which were noise.
Stopped pulling data by hand. Built a Python pipeline to fetch it automatically, then used R to work out which factors were actually driving returns and which were noise.
Wrote a Python program pulling daily price and macroeconomic data via API (Yahoo Finance, Federal Reserve FRED), removing manual collection from the workflow entirely. Ran and interpreted OLS regressions in R to quantify sensitivity to equity, oil, currency, and Treasury-yield factors.
Isolated the statistically meaningful drivers from the noise and documented the finding for a non-technical reader. The automation half is the part I reuse constantly — the analysis is only as fast as the data getting to you.
Two public companies, multi-year SEC filings, no shortcuts. Built the full analysis from primary documents and traced the performance gap to the specific line items causing it.
Two public companies, multi-year SEC filings, no shortcuts. Built the full analysis from primary documents and traced the performance gap to the specific line items causing it.
Constructed ratio, common-size, cash-flow, and net-working-capital analyses in Excel using PivotTables and VLOOKUP, plus forward-looking pro forma projections. Worked from the 10-Ks directly rather than summarized data, examining the accounting treatments behind the reported numbers.
Delivered a written recommendation identifying the stronger financial position and the exact reported line items driving the difference. The habit it built — never trusting a figure I haven't traced to its filing — is the one I apply to every model since, including anything AI touches.
Selected for a live-client operations case. Analyzed patient-flow and capacity constraints at a major health system and quantified what fixing them was worth.
Selected for a live-client operations case. Analyzed patient-flow and capacity constraints at a major health system and quantified what fixing them was worth.
Evaluated the operational bottlenecks limiting throughput, then applied scenario and return-on-investment analysis to size the trade-offs between competing improvement options.
Presented recommended operational improvements alongside the process and capital changes required to execute them — not just what to fix, but what it would cost and return.
I don't just use AI — I build with it and I check its work. Deployed an agent that runs unsupervised against a live account, and I'm building an app that scores how deeply AI is integrated into a workflow. In financial analysis, AI speeds up my first pass; verification decides what survives.
I don't just use AI — I build with it and I check its work. Deployed an agent that runs unsupervised against a live account, and I'm building an app that scores how deeply AI is integrated into a workflow. In financial analysis, AI speeds up my first pass; verification decides what survives.
Built and deployed a Gmail agent that autonomously classifies and clears promotional email, with a deliberate safeguard — flagged mail routes to Trash on a 30-day recovery window, never permanent deletion. Currently developing an AI-readiness scoring application, and apply AI across modeling and research to accelerate first-pass analysis.
The agent runs unsupervised in production. In financial work, tracing every AI-generated figure back to source has already caught calculations that were wrong in ways that looked entirely plausible. The real result: speed I can rely on, and a verification habit I now apply to every model I build.