Devam Chaturvedi

AI Product Manager with a Computer Science engineering foundation and a Postgraduate Diploma in Applied Statistics (Indian Statistical Institute), who owns products end to end — from discovery through architecture to launch. Solo-designed, built, and launched PrepParity, an AI-assisted SaaS exam-prep platform: architected a 124-endpoint API on Next.js/FastAPI/PostgreSQL and a verbatim-retrieval AI design built to prevent hallucination by architecture rather than prompt tuning, validated against a 2,551-question regression corpus. Two-plus years earlier translating enterprise stakeholder requirements into shipped features at Publicis Sapient.

Brings technical depth, statistical grounding, and hands-on AI product judgment to work directly with engineering and ML teams on system design, evaluation, and tradeoffs.

Founder & Product Lead at PrepParity (2025-12 – Present)

Own the full product lifecycle for PrepParity, an AI-assisted SaaS exam-preparation platform for UPSC Civil Services aspirants, from problem definition through architecture, build, launch, and iteration. Launched approximately June 2026.

• Conducted 10–12 discovery interviews with UPSC aspirants sourced from online and offline aspirant communities to validate problem framing before committing to the first feature set.
• Defined product strategy, roadmap, and a prioritized backlog with written user stories and acceptance criteria; ran release planning independently across a ~7-month build (December 2025 to launch, June 2026).
• Designed the product's core AI architecture around verbatim source-text retrieval rather than content generation, built to prevent hallucination by architecture rather than relying on prompt tuning — critical in a product where a wrong answer costs a student marks.
• Built and validated a 2,551-question content-accuracy regression corpus (1,499 Prelims, 1,052 Mains) spanning the full GS1–GS4 and Essay syllabus, tracked against 2,699 syllabus topics.
• Specified the learning engine — active-recall enforcement and SM-2 spaced repetition — translating published learning-science research into concrete acceptance criteria.
• Architected and built the full stack on Next.js, FastAPI, and PostgreSQL using an AI-assisted development workflow: 124 versioned, authentication-gated API endpoints, 73 row-level-security policies, and 7 idempotent background jobs, across roughly 22.6K lines of backend Python (93 files) and 111 frontend files.
• Wrote data-privacy and access-control requirements into the specification ahead of build, and load-tested the platform to 1,000 concurrent users.
• Directed a 3–4 person external marketing team and coordinated go-to-market messaging with the product roadmap.
• Validated paid-channel economics with a ~$71 Reddit acquisition test (125,122 impressions, 1,983 clicks, 1.59% CTR, ~$0.04 CPC) ahead of scaling spend.
• Drove the platform to 200+ registered accounts, 2.63K visitors, and 3.12K sessions in the first month post-launch (5m 47s avg. session, 38% bounce rate, ~72% mobile traffic), tracked through product analytics.
• Secured DPIIT Startup India recognition for the company, managing the regulatory filing independently.

Associate Technology L2 at Publicis Sapient (Publicis Groupe) (2018-06 – 2020-08)

Approximately 2+ years including technology traineeship (January – June 2018). Led client-facing requirements gathering across 4 enterprise retail projects, translating stakeholder needs into technical specifications, feature scope, and delivery requirements.

• Delivered end-to-end digital storefront features for Dolce & Gabbana, Fresh, L Brands, and Shiseido — from solution design through front-end components, backend integrations, and release — on Salesforce Commerce environments.
• Worked within 10–12-person Agile teams, prioritizing backlog items across sprint planning, refinement, demos, retrospectives, and releases, using Jira and Confluence.
• Applied data analysis and technical understanding to inform client decisions and product direction across storefront features.

Independent Quantitative & Policy Research | UPSC Civil Services Examination at Self-Directed (2020 – 2024)

Full-time candidate for the UPSC Civil Services Examination: India's national civil-service entrance exam, with a selection rate under 0.5% among ~1M+ annual applicants — cleared the Preliminary stage and reached the Mains stage twice across four years of self-directed, full-time study under exam conditions.

• Researched and synthesized dense material across governance, economics, history, and current affairs into structured briefs; wrote timed, structured responses under strict word and time limits across nine descriptive Mains papers.
• Built quantitative reasoning and structured problem-solving through the exam's CSAT component, and ran a self-directed multi-year study program — prioritizing topics and reallocating effort based on mock-test performance across four Prelims attempts and two Mains qualifications.

Post Graduate Diploma in Applied Statistics in Data Analytics – Indian Statistical Institute (ISI) (2025-02 – 2026-03)

B.Tech in Computer Science & Engineering – Graphic Era University (2014 – 2018)

Global MBA (Online) – University of Western Australia (2026-03 – 2027-03)

Advanced Business Management Programme – Indian Institute of Management (IIM) Kozhikode (2024-12 – 2025-12)