TEJAS VENKATESH TEJAS VENKATESH

Evolved a single-domain BESS tracker into an AI-powered market intelligence platform spanning 8 industries, auto-categorizing articles into competitor, innovation, and policy buckets — each tagged with an impact rating to surface what actually matters.

Runs fully automatically, continuously scanning sources so every industry digest stays up to date without manual intervention.

Engineered a multi-provider LLM pipeline (Groq, OpenRouter, Cerebras) with fallback handling for rate limits and API errors, keeping the digest running reliably on free-tier infrastructure.

Added an "Ask AI" feature letting users query the corpus conversationally instead of manually combing through sources.

Built an AI-powered tool that turns a pile of customer reviews into decisions — scoring sentiment, surfacing recurring issues, and producing a ranked action queue with a suggested reply for every review.

Auto-tags recurring problems across quality, shipping, sizing, safety, pricing, and service, then ranks the reviews that need a response first so teams tackle what matters most.

Added drill-down analytics — click any KPI to see the reviews behind it, plus word clouds and sentiment trends.

Ships with a realistic 180-review sample and supports CSV upload, replacing manual spreadsheet triage and hours of reading reviews by hand.

Built an interactive way to learn system design by taking real architectures apart — open a proven design, click any component to see why it was built that way, then swap a piece and watch what breaks.

Explains trade-offs in plain language — the same ones an interviewer actually asks about — with every explanation hand-written and fact-checked, not guessed by an AI.

Ships opinionated reference designs across domains — ticket & slot booking, social media feeds, e-commerce checkout, and live betting exchanges — each showing the downstream cascade of a single decision.

Includes a quiz mode (predict before you reveal), back-of-envelope estimation, and cross-domain patterns that transfer between problems.

Designed a high-fidelity Figma prototype for a ticket-booking app spanning multiple event types, taking users from discovery all the way through checkout.

Built 4 complete end-to-end workflows — browsing, event details, ticket selection, and payment — fully clickable and interactive.

Used carousels, micro-interactions, and varied animations to create a fluid, engaging booking experience.

Crafted a polished UI/UX with a cohesive color system and clear visual hierarchy that keeps every flow intuitive.

Built an interactive Tableau dashboard consolidating siloed EV datasets (EPA fuel economy, state incentives, charging infrastructure) into a single visualization platform for consumers, policymakers, and manufacturers.

Surfaced CAFV eligibility insights showing 68% of analyzed EVs qualified for clean-fuel incentive status, alongside an average emissions reduction of 4.6 metric tons CO₂ per vehicle annually.

Designed geographic heat maps, time-series adoption trends, and interactive filters that cut EV buyer research time from hours to minutes.

Identified a citywide parking pain point: drivers spend 30+ minutes daily searching, fueling congestion and productivity loss.

Designed a scalable IoT parking platform with sensors, gateways, and cloud for real-time spot detection and AI routing for 2- and 4-wheelers.

Produced detailed product specs and a full Figma prototype covering reservation, payment, and navigation flows.

Established a transaction-led business model targeting a 2% take rate.

Conducted comprehensive user research on visually impaired market, identifying 76% loneliness rate in visually impaired individuals vs 25% in general population, representing 2.2 billion potential users

Developed innovative product strategy addressing critical accessibility gaps in social connection platforms, defining 5+ key features and requirements for this underserved user segment