Interview with Nav Deol MBA, Advisor, Massachusetts Institute Of Technology / Westgate

This interview is with Nav Deol MBA, Advisor, Massachusetts Institute Of Technology / Westgate.

To introduce yourself to CFO Drive readers, how do you describe your work today advising technology ventures?

I’m passionate about advising technology ventures and find it very fulfilling work. From a young age, I’ve been interested in technology, and helping founders and startups succeed allows me to make substantial, positive contributions to the field.

I typically advise startups on fundraising, strategy, and positioning to investors. I especially enjoy bridging the gap between technical founders and investors, as it requires translating a founder’s vision and technology into business goals and hard numbers.

What key decisions or experiences led you from finance and Investing into advising for early-stage startups?

While working at large financial institutions, I often advised corporations on all aspects of business. I've worked across the capital structure and assisted with transactions ranging from structured debt to equity raises. Transitioning to work with early-stage startups therefore seemed like the natural next step.

I find that I can add even more value to startups that don't have the manpower of a full finance division behind them. At startups, I mostly work with highly technical co-founders who need help bridging the business side of things.

Drawing on your AI and analytics focus, what practical framework do you use to embed AI into underwriting and portfolio decisions?

In my estimation, AI integration into underwriting and portfolios is about creating an ongoing process rather than relying on a one-time model.

It begins with using better inputs—leveraging AI to mine decision-relevant information from operations and behavior, along with financial data.
It continues by helping underwrite through scenarios and probabilities, always with the involvement of human expertise, particularly in complex situations where AI needs better context and human experience.
The real added value comes after origination by informing decisions on positioning, pricing, and monitoring, as well as by creating a mechanism for reassessing decisions when circumstances change.
The cycle is then closed and made more effective by comparing results and learning from both models and decision-making processes.

Building on that, what lightweight analytics stack do you stand up in the first 30 days to generate decision-grade insights for a finance team?

The first month is all about speed rather than elegance. I generally build a basic database layer using something like Snowflake or BigQuery, ingest core finance and operations data, and get things cleaned up with dbt so I can start producing metrics quickly.

Database layer: Snowflake or BigQuery
Data transformation: dbt to clean and model data
Business intelligence: Looker for real-time visibility
AI layer: an AI provider (for example, OpenAI) for lightweight scenario analysis and anomaly detection

The point at 30 days is not to achieve perfection but to establish the one single source of truth.

Shifting to origination, how do you design a differentiated deal-sourcing engine that blends data pipelines with venture and corporate networks?

A differentiated deal-sourcing engine is much more about timing and access than scale. From my personal experience, the key is marrying a data-driven sense of companies' early inflection points with tightly networked venture and corporate pipelines.

For data, you triangulate signals that exist prior to a company going viral, so looking at:

hiring trends
changes in product adoption
shifts in financing
ownership

These signals alone are not very valuable, but taken together they often reveal inflections that point toward breakout companies. Insights are only valuable when action follows, and that action comes through the management of key networks.

From experience in building deals across the capital stack, I have seen that promising ventures will flow through founder, operator, and investor relationships long before hitting the mass market. Managing your pipeline of contacts is a prerequisite to accessing truly valuable opportunities.

Together, the melding of data and networks creates your differentiated deal-sourcing engine.

When diligencing GenAI or frontier-tech startups, how do you evaluate technical teams and defensibility based on your MIT and founder experience?

In conducting diligence on GenAI or frontier-technology startups, my background at MIT and my work with founders has led me to look beyond teams that can merely assemble available technologies and instead to look deeper for teams that understand how their models behave, where they are vulnerable, and the trade-offs of their design.

One indicator of a successful startup is a team that understands its product or model in enough detail to explain its functionality and how it will evolve through further iterations. This becomes an even greater advantage in the current environment of model commoditization.

Defensibility is not in the models themselves; it lies in how the compound effects of the model are achieved:

Proprietary data
Constant iteration with the help of customers
Improvements driven by workflows

For alternative and private equity deals, what early red flags in models or operating metrics most often change your investment thesis?

In alternative investments and private equity transactions, the first indications of problems are often where the story conflicts with the economics.

Rapid growth in revenues while unit economics are poor or deteriorating, especially if margins, CAC, or payback periods don't show improvement.
Unrealistic expectations about growth, pricing, or exit multiples that are not supported by the operations.
Cash generation as a screening tool: a lack of cash conversion from earnings, or reliance on working capital growth to generate growth, almost always signifies underlying difficulties.
On the operations side, inconsistency - in churn rates, cohort performance, or concentration ratios - often indicates hidden weakness.

Any disconnect between the story and the economic fundamentals triggers a rethinking of my investment thesis.

After an investment closes, how do you partner with CFOs to drive EBITDA and cash improvements that show up within two quarters?

Following the close, we work with the CFO on several key movers of EBITDA and cash in the short term. This often starts by establishing a clear baseline, followed by rapid actions around pricing, cost containment, and working capital management.

Small pricing adjustments or discounts can drive quick margin improvements.
Cost savings from tighter expense controls and vendor relationships result in EBITDA improvements.
On the cash side, accelerating collections, paying vendors later, and trimming inventory levels typically flow through in 90 days or less.

It's critical in these early days to maintain a tight weekly rhythm around EBITDA and cash metrics.

For CFOs starting small, what 90-day pilot would you run to introduce AI into FP&A or corporate development?

For a new CFO, the first 90 days should include running a targeted pilot for a single workflow, i.e., forecasting or deal review, where AI could prove its worth.

Examples:

In FP&A, you could build a simple model that takes historical financials and drivers and builds a rolling forecast, as well as variance analysis explaining changes.
In corporate development, you can build a pipeline system that analyzes companies and reviews the content, then starts flagging firms based on your preset criteria for investments.

You should use available tools and existing datasets rather than build brand new infrastructure.

In under 90 days, the goal is to achieve results that don’t interfere with core operations but add value (through improved output or faster processes).

If the project is successful, move on to applying AI to another workflow.