Building Solstis

• Developed investment strategies for 99Tartans' new Emerging Markets division, focusing on connecting CMU alumni startups with CMU investors. • Built a network of investors in US and Emerging Markets, supporting two startups with fundraising efforts, while being one of only two non-MBA partners in a predominantly MBA team. • Conducted due diligence on early-stage CMU-affiliated startups, assessing market potential, competition, and business viability. • Hosted the first open-mic pitch night featuring renowned founders from India as panel judges, fostering connections and showcasing innovative CMU-affiliated ventures.

Member of 99 Tartans, an angel investment syndicate of Carnegie Mellon alumni that invest in alumni-founded startups. Act as a first point-of-contact for early-stage startups seeking pre-seed, seed, and Series A funding.

• Organized town-hall style feedback sessions to gather student input, directly contributing to curriculum adjustments and program enhancements tailored to student needs. • Acted as a key liaison between the engineering department and students, facilitating collaboration and ensuring that students’ voices are heard by faculty and administration. • Helped Mechanical Engineering (MechE) professors define and refine the structure of specific programs within the department, ensuring alignment with both academic standards and student aspirations.

Explored industry-relevant solutions by adapting a diffusion-based approach from an ICRA paper for real time 3D pose estimation, focusing on mitigating inaccuracies caused by occlusion for a construction site. The project was sponsored by YKK AP

• Built a framework to automate the creation of dataset schema mappings for synthetic healthcare databases within unity catalog, reducing manual mapping by 60%. • Designed a custom pipeline for Retrieval Augmented Generation (RAG) using Databricks embedding models, integrating FAISS as vector store, while leveraging both Azure OpenAI and DBRX. • Established a structured approach utilizing Mosaic AI’s agent framework to construct HIPAA-compliant RAG applications to reduce setup time by 70%. • Created a base architecture of Agents and Tools via the LangChain framework to demonstrate Gen AI's strength in descriptive tasks and discovery analytics as a proof of concept.

B2C entrepreneurial AI powered platform that aimed to guide early stage entrepreneurs through the process of converting their ideas from a POC to a business model by consolidating and visualizing all of their resources into one platform. Talked to 50+ entrepreneurs and pitched this idea in a AI Startup course to a visiting VC. Ultimately paused because we couldn't reach problem validation and pivoted to finding a different problem.

• Developed and executed test scripts in CANape/CANoe, leveraging Vector tools to perform durability tests on test benches, resulting in a 10% reduction in downtime. • Automated logging processes on various test benches, improving data accuracy by 20%. • Designed, developed, and resolved problems of new and existing software, HIL models, test scripts and components of powertrain products for the new Electrification platform. • Lead and supported DVP and DVPR, identifying component, system, and vehicle level testing required during the development phase. I worked on multiple projects while also doing a stretch assignments with other teams.

• Developed and performed hardware-in-loop (HIL) simulation and testing by developing real-time sequences and layouts through Automation Desk and Control Desk respectively. • Used MATLAB & Simulink to develop energy management strategies and fault detection for efficient and safe operation. • Developed requirements and tested all control code iterations in MIL & HIL simulation environments in addition to extensive in-vehicle testing.• Generated and analyzed various signals and CAN messages to develop, test, and validate the code running the hybrid powertrain systems on a modified Chevy Blazer.• Implemented reliable and robust vehicle powertrain control algorithms using an iterative development process and closed-loop testing while gaining competencies working within MATLAB and Simulink.

• Explored the various aspects of automotive communications through the lens of cybersecurity.• Interacted with CAN, LIN, ECUs, OBD II ports, and wireless interfaces.• Evaluated/ inspected message traffic across CAN and future migration to automated vehicles.• Segmentation of control functions within a smart car, including external controls, infotainment systems, and passenger interaction.

- R1T, R1S, and Amazon EV (RPV) product launches

Designed a CAD model of an autonomous robot in Siemens NX. Evaluated various electrical components like LIDAR and real time imaging by combining ROS and MATLAB and rapidly tested the working of the autonomous system of the robot.

• Completed weekly destructive test to check for weld defects by completly destructing the back panel (ADA-Elantra) • Inscribed over holes on parts using part drawings by placing them on fixtures, understanding tolerances and acceptance criteria (Press part defects) • Worked on part rejection and identification procedure to check whether a part needs to be scrapped or sent to rework • Made audit reports, attended and effectively contributed to the “eagle eye process” that was held everyday• Reduced defects such as overshadow and weld spatter by introducing and implementing Bolt Checking process (coined the term “Bolt Audit”) during the eagle eye process • Performed thickness testing on parts of the new Hyundai Sonata using Ultrasonic Thickness Tester and used inspection tools like Vernier Calliper, Scales, Flush Gauge, Ultrasonic Weld Testing Machine, and Universal Testing Machine • Completed the assigned responsibility of measuring 4 critical points on P/Tray (ADA-Elantra) using Flush Gauge, palletizing all the samples and shipping of these pallets to the customer assembly line ensuring Just in Time (JIT)