AI Product Engineer

SARC Global | Delhi NCR | Full time | Reports to the CEO

About SARC

SARC is a 40-year-old multidisciplinary advisory firm with 100+ partners and 500+ professionals across India, UK, UAE, Singapore and the USA. We advise India's regulated enterprises — banks, NBFCs, insurers, PSUs on cybersecurity, data protection (DPDP), and regulatory compliance. We are now integrating that advisory depth into products: compliance platforms, assessment tools, and AI-native internal systems.

The Role

This is a one-person product team sitting in the founder's office. You will take what our consultants do - DPDP compliance, cyber risk assessments, compliance tracking and turn it into robust forward customer centric applications. You are both the engineer who ships and the PM who decides what to ship.

You are hands on with AI tools and open-source is your default stack. You will work directly with the CEO and with practice leaders (cyber, data protection, tax, regulatory) who are your users and domain experts.

What you will build

- DPDP compliance modules — data discovery workflows, RoPA builders, consent tracking, gap-assessment engines that turn our advisory methodology into a repeatable product
- Cyber assessment tooling — questionnaire-to-report pipelines, control-framework crosswalks (NIST CSF, ISO 27001, RBI, SEBI CSCRF, CERT-In), evidence collection and scoring platforms that productise how we deliver assessments
- Existing platforms — extend and harden our in-house products (compliance automation, tax notice management) built on Next.js, Supabase/Postgres, and the Claude API
- Internal platforms and AI workflows — proposal generators, document intelligence, research agents that make 500 consultants faster

What you will do

- Own products end-to-end: talk to practice leaders, scope the MVP, build it, ship it, iterate on real usage
- Build fast with Claude Code, LLM APIs, and open-source components - assemble before you build, build only what differentiates
- Design pragmatic architectures: RAG pipelines, structured LLM outputs,



agentic workflows, evals — applied to compliance and cyber use cases
- Run the infrastructure you ship (AWS EC2, Supabase, Postgres, PM2/nginx, CI) — no separate DevOps team
- Write just enough documentation that the firm isn't dependent on your memory
- Say no. Prioritise ruthlessly across many stakeholders with a one-person team

Must Haves

- 2–6 years building and shipping full-stack products, ideally solo or in very small teams
- Strong TypeScript/JavaScript (Next.js/React) and working Python; comfortable with Postgres and SQL
- Hands-on with LLMs in production: prompt design, RAG, structured outputs, tool use, basic evals not just API calls in a notebook
- Daily driver of AI coding tools (Claude Code, Cursor, or similar) - you ship 5–10x faster than a traditional developer and can prove it
- A strong portfolio of complex applications you've actually shipped: GitHub, side projects, internal tools, indie products. We weigh this over pedigree
- Product judgement: you can sit with a non-technical partner, extract the real requirement, and cut scope to a shippable v1

Nice to Have

- Exposure to BFSI, compliance, GRC, or audit domains (DPDP, RBI, SEBI, ISO 27001)
- Strapi or headless CMS experience; AWS operations; Playwright/scraping
- Open-weight model experience (vLLM/Ollama, fine-tuning) for sovereign-deployment scenarios
- Prior founder, founding-engineer, or agency experience where you owned outcomes, not tickets

Who thrives here Someone who finds a 40-year-old firm full of unsolved, monetisable problems exciting rather than unglamorous. You'd rather ship four working tools this quarter than perfect one. You're comfortable being the only technical person in the room and translating between regulators' language and code.

This is not for you if you need a spec handed to you, a team to delegate to, or a pure-research ML role.

How to Apply

Send your resume plus links to 2–3 things you've built (repos, live tools, demos) and 3–4 lines on which of our problem areas you'd build first and how. Applications without shipped work will not be prioritised to [email protected].

📌 AI Product Engineer (Engineer + PM) (New Delhi)
🏢 sarc
📍 New Delhi