Boston University '26 · CFA Level I
Ritwik
Saini
Private Capital Markets. I source deals, build theses, model investments, and build automations around financial workflows, sitting at the intersection of rigorous financial analysis and technology-driven execution.
I bring systems-level thinking to private markets: breaking problems down to first principles, whether that's a deal thesis, due diligence, a sourcing process, or a workflow that shouldn't need a human in the loop.
That's the thread across venture, private equity, and the automations I build around both: understanding a business by its structure, not its narrative. I'm particularly drawn to energy-tech, deep-tech, and AI infrastructure, sectors where the technology itself is the thesis.
Outside of that: I play drums, spend a lot of weekends hiking, and listen to more rock music than is probably useful for anything.
Limbic Capital
Private Equity Intern · London, UK
- ·Built a sourcing infrastructure from scratch: automated outreach across M&A Advisory, Manufacturing, Real Estate, and HVAC, generating 20 first meetings and 3 signed NDAs in the first month. Sourcing as a system, not a hunt.
- ·Mapped 500+ companies and 370+ contacts across RNG/AD developers and European lower mid-market M&A advisors; built a proprietary database of 7,000+ US companies tracking M&A activity to give the team a structural coverage advantage.
- ·Red-teamed the data center investment thesis from first principles: probability-weighted scenario models, primary hard signals, kill-criteria stress testing. Conclusion: mode selection, not sector timing, determines whether the thesis works.
- ·Created the end-to-end marketing pipeline of an alpha-stage portfolio SaaS company and deployed an analytics layer that autonomously corrects copy when metrics breach floor thresholds, closing the feedback loop no one had automated.
Eaglewings Ventures
Investment Analyst Intern · Mumbai, India
- ·Developed investment theses on Nuclear Energy in India and Immersion Cooling, surfacing nuclear as India's cheapest firm power at ₹3.83/kWh vs. ₹5+/kWh for firmed renewables and expanding the pipeline with 3 qualified companies.
- ·Supported pre-Series A diligence across 15+ founder calls (burn rates, cap tables, unit economics), building evaluative frameworks that outlasted the specific deals.
- ·Produced 2 sector research reports mapping regulatory risk, investment potential, and scalability across DeepTech and energy verticals.
Caprae Capital Partners
Private Equity Intern · Glendale, CA
- ·Built 2 LBO models across Legal Tech and ITAD targets, translating financial diligence into go/no-go acquisition memos with clear investment rationale.
- ·Sourced acquisition targets through direct founder outreach: cold-called into 2 partner-level meetings, proving that disciplined prospecting beats warm introductions when the thesis is specific enough.
- ·Produced 3 sector reports across Legal Tech, ITAD, and Accounting, mapping market structure, competitive dynamics, and where acquisition value concentrates in each vertical.
HP Tech Ventures
Deal Sourcing & Startup Analysis Extern · Remote
- ·Screened 20+ startups against HP's investment theses, surfacing 5 high-conviction targets for deeper review; one advanced to partner-level due diligence. Selection criterion: teams building infrastructure layers, not features.
The Dallas Entrepreneur Center Network
Project Management Intern · Dallas, TX
- ·Replaced intuition-based mentor-mentee pairing with a structured, metrics-driven dashboard, driving a 24.8% improvement in match quality. Good system design beats good intentions.
- ·Coordinated delivery across a team of 5, maintaining 95% on-time project completion by building the tracking infrastructure that made the rate achievable.
Udayaa
Lead Software Engineer · Remote
- ·Led architecture and delivery of a full-stack platform for a non-profit supporting student fundraisers, directing a 5-person team from spec to live product using React, Django, PostgreSQL, Docker, and AWS.
Sector theses built from first principles. The thesis, the data, and a recommendation.
Eaglewings Ventures · 2025
Nuclear Energy in India
"Nuclear is India's cheapest firm power at ₹3.83/kWh, decisively below firmed renewables at ₹5+/kWh once storage costs are factored in."
Key Findings
- ◆TAM expanding from $1.54B today to $100B+ by 2047, at an 11% CAGR vs. global 2.9%, driven by peak load demand tripling to 366.4 GW by 2032.
- ◆$2.4B government allocation in the 2025 budget targets 5 operational SMRs by 2033 and $26B in projected private investment. Atomic Energy Act reforms opening up to 49% FDI.
- ◆AI-driven data center growth and 20-year hyperscaler PPAs (Microsoft, Google, Amazon) are creating durable structural demand for firm, dispatchable baseload power.
Recommendation
Express via EPC contractors and fuel-cycle suppliers rather than direct generation assets. Monitor Hylenr (LENR) and Pranos Fusion (spherical tokamak) as early-stage private plays with asymmetric upside.
Limbic Capital · 2026
Data Center Investment Thesis: Red-Team
"The sector thesis survives, but mode selection determines returns. A build-led platform creates a ~32% development margin; acquisition-led roll-ups at 28x EBITDA are a leveraged bet on peak multiples persisting, not a manufactured spread."
Key Findings
- ◆$600B AI capex-to-revenue gap is widening; Microsoft cancelled 200 MW of US leases; Virginia Digital Gateway (the largest campus ever proposed) was killed by permitting opposition. A 35% combined probability of negative equity returns across bear/break scenarios requires only modest, correlated deterioration.
- ◆+200bps cap rate expansion erases the entire value-creation margin. The 2028–2030 refinancing wall concentrates this rate risk into a narrow window.
- ◆Power is simultaneously the thesis's strongest moat (3–5 year grid interconnection queues) and its most acute tail risk: 75 projects worth $130B were delayed or blocked in Q1 2026 alone.
Recommendation
The only version worth backing is build-led, pre-leased to investment-grade hyperscalers (50–70%), and power-secured before a dollar of hard cost is committed. Anything else (acquisition-led at 25–30x, speculative development, neocloud tenants) is equity risk wearing a real estate costume. Refuse above 20–22x EV/EBITDA. Actively manage the 2028–2030 refinancing wall; at +200bps the entire value-creation margin disappears.
Python · Claude API · FastAPI · React · Vite
- ◆PE deal triage follows the same pattern every time: ingest documents, extract the metrics that matter, write the memo. I automated all three with a multi-agent LLM pipeline, reducing analyst write-up time from days to minutes.
- ◆60% faster on metric extraction. An additional 30% efficiency gain from integrated LBO Modeling and Research agents that feed conclusions directly into the IC Memo. The model does the grunt work; the analyst does the judgment.
Python · Django · React · PostgreSQL · Docker · AWS
- ◆BU's course planning system forced students to manually check 7,000+ courses against graduation requirements, a coordination problem cleanly solvable by a greedy algorithm. Built the algorithm, then the product. 1,500+ students used it.
Finance & Investment
Tools & Data
Sep 2022 – May 2026
Boston University
BA Computer Science & Economics
Aug 2020 - April 2022
Pace Junior Science College
HSC (Class XII) · Science Stream
Let's Talk
Open to VC, PE & IB Opportunities
and discussing energy-tech, AI Infra and deep-tech.
I'm pursuing full-time roles in PE, VC, and IB. If you're looking for someone who can source deals, build investment theses, and automate financial workflows, let's connect. I'm also happy to chat about energy-tech, AI Infra, and deep-tech.