AI Engineer

Location: Chandigarh / Bangalore

Job Type: Full-Time (Hybrid Mode)

Sprouts.AI is an AI native “Generative Demand Platform” disrupting the $1.1T B2B GTM category. It helps B2B demand gen and sales teams generate qualified pipeline and increase deal conversion rates with real-time account and contact level intelligence signals, recommending next best action and execution via AI agents. We are solving the problem of inefficient GTM driven by dirty data ecosystems, multiple siloed legacy tools, large GTM teams, static ICPs and poorly defined segments. Current customers have seen a significant improvement in data coverage and quality, 3X increase in qualified leads, over 50% increase in SDR/BDR efficiency, and a 30%+ reduction in martech tool cost.

We’re looking for exceptional developers who are excited to build in an AI-first way: coding with AI (Cursor/Copilot/agentic coding), shipping at speed, and wiring real GTM data into decision-making agents. If you want to own end-to-end features and move fast with AI as your co‑developer, this is for you.

What you’ll own (outcomes, not tasks)-

• Ship agentic features end-to-end: from a bare prompt to a production feature (UI + backend + agents + evals).
• Operationalize RAG/agents: retrieval quality, tool-use reliability, guardrails, and observability (traces, feedback loops).
• Integrate GTM data sources: enrichments, scoring, dedupe, and live syncs across CRMs, enrichment APIs, comms tools.
• Design for speed: scaffold with AI, iterate in days (not weeks), demo early/often, measure usage and impact.
• Harden what sticks: once a prototype hits fit, add tests, typed interfaces, retries, and idempotency—without killing velocity.

Tracks you can lean into-

• Frontend-heavy: Next.js/React/Tailwind; conversational/UI for agents, copilots, evaluators, dashboards.
• Backend-heavy: Python/Node APIs, agent orchestration, data pipelines, vector stores, eventing.
• Agentic orchestration: LangGraph/LangChain/LlamaIndex, tools/functions, RAG, evaluators, experiment harnesses.
• Full stack: Capable of working across frontend and backend, integrating agents, APIs, and data flows into cohesive product slices; comfortable switching contexts and delivering E2E features at speed.

(You don’t need to be great at all four—be excellent at one and competent at the others.)

Tech you’ll touch

• AI & orchestration: LangGraph, LangChain, LlamaIndex; function/tool calling; evaluators; tracing.
• LLM ecosystem: OpenAI/Mistral/Gemini, embeddings, vLLM/server choices, prompt/version control.
• Data & retrieval: Postgres/pgvector, Milvus/Weaviate/Chroma, Elasticsearch, Redis, basic Kafka/eventing.
• Web stack: TypeScript, Next.js/React, Node.js/Express, Tailwind; Vercel/Fly/Docker; CI/CD.

Must-have skills (hard filters)

• Excellent coding skills: Strong command of Python and/or JavaScript/TypeScript; ability to write clean, maintainable, scalable, and efficient code across backend/frontend.
• AI-assisted coding as default: you already use Cursor/Copilot (or similar) daily and can show it in action.
• LangGraph (or its alternatives) proficiency; retrieval evaluation experience; E2E analytics (events, funnels); background jobs & queues;
• Shipping DNA: portfolio of small products/prototypes shipped fast (hackathons, indie projects, OSS, client work).
• Agent pragmatism: you’ve built at least one real agentic thing (even scrappy) that calls tools/APIs and returns value.
• Systems thinking: can sketch a flow from data → retrieval → reasoning → action → feedback loop.
• Quality under speed: basic tests, typed boundaries, retries/backoffs, and logging without ceremony.
• Prompting & communication skills: Ability to design effective prompts, collaborate through pair-prompting, and clearly communicate ideas and trade-offs with both humans and AI models.

Nice-to-haves (signal boosters)

• Real integrations with any CRM/enrichment/outreach tool; design sense for AI UX (structured inputs, affordances).
• Exposure to LLMOps tools (e.g., evaluation frameworks, tracing/monitoring platforms).
• Contributions to OSS, hackathons, or side projects that showcase fast prototyping with AI.

You’ll thrive here if you…

• Reach for AI before boilerplate; pair-prompt with teammates; write small design notes and iterate.
• Demo half-working things to get feedback; love “thin slice → ship → instrument → refine.”
• Treat data contracts/observability and eval based development as part of product quality, not “later.”

Please don’t apply if…

• You want long upfront design cycles and weekly status over frequent demos.
• You avoid AI coding tools or see them as “cheating.”
• You only want research/papers, not product shipping.

Hiring signals we look for

• Repos or demos of agentic and AI native features (even rough).
• A story where you shipped in 48–72 hours with AI tools.
• Clearheaded trade-offs (latency vs. quality, retrieval vs. reasoning, eval cost vs. benefit).

Anti-signals

• “I’ll build it all by hand first and automate later.”
• Over-indexing on framework dogma; refusal to instrument/evaluate.
• “Agents are magic” (no plan for tools, states, or recovery).

What you’ll get

• Competitive salary and meaningful responsibility from day one in a high-trust, builder-led environment.
• Fast path to broader ownership and leadership for candidates who consistently ship and raise the bar.
• Direct collaboration with world-class founders, technologists, and AI leaders—exposure to cutting-edge thinking and mentorship.
• A chance to work on frontier problems in GTM and AI-native product design, where your work has immediate impact.
• Access to top-tier AI tooling, workflows, and a culture of experimentation and fast learning.
• An environment where demos, learning, and iteration are valued over bureaucracy and slow cycles.