AI Engineer— AI-First Product Builder
Worksmith Solutions | Full-Time | Remote / Hybrid
About Worksmith Solutions
Worksmith is a fast-growing AI consulting, growth and technology firm helping clients build scalable systems, automate operations, and grow through technology. We move fast, build smart, and leverage cutting-edge AI tools to deliver work that would take traditional agencies 3x longer. We're all about speed, accuracy, and perfection!
The Role
We're looking for a Tech Lead who builds with AI at the center, not as an afterthought. You'll be responsible for architecting and shipping full-stack digital products — from CRM systems and patient portals to automation pipelines and mobile apps — using AI (ideally Claude by Anthropic) as a core productivity multiplier across the entire development lifecycle.
This is not a role for someone who uses AI occasionally. We want someone who thinks in prompts, designs agentic systems the way others design databases, ships in sprints, and solves problems before they become tickets.
Our website — worksmith.tech
What You'll Build
Based on our active client roadmap, you'll be delivering:
Business Health Check — a financial and operational diagnostic for businesses: automated reconciliation and leakage detection across procurement, site cash control, labour, and collections, delivered as a measured, evidenced finding
CA Workflow Automation — practice automation for Chartered Accountant firms, priced per entity under management: GSTR-2B reconciliation, ITC risk monitoring, and Tally/Zoho-integrated workflows across a firm's full client book
Operational Intelligence Agents — the multi-agent system underpinning both products above, including agents that write back into a client's live accounting system under guardrails, not just read and report
AI-Powered Internal Tools — Claude-driven agents for content generation, reporting, and internal workflow automation
Core Responsibilities
CRM Systems — custom-built data systems with referral tracking, follow-up automation, and dashboards
Web Portals — patient portals, doctor portals, and corporate health dashboards
Mobile Applications — cross-platform apps for appointment booking, records access, and push notifications
Marketing & Lead Gen Tech — landing pages, WhatsApp automation pipelines, lead funnels
SEO & Analytics Infrastructure — structured data, performance tracking, conversion dashboards
WhatsApp & Email Automation — multi-step nurturing sequences and consultation scheduling bots
Lead end-to-end technical architecture and delivery of client projects
Use Claude (Anthropic API) as a primary tool for: code generation, debugging, documentation, content automation, prompt-driven UI scaffolding, and internal tooling (including your programming knowledge and edits wherever needed)
Design, evaluate, and iterate on prompt architectures, agentic workflows, and RAG pipelines for internal and client-facing AI systems
Own the reconciliation and leakage-detection engine behind the Business Health Check, and the GST/entity-based automation behind the CA Workflow product — these are the two flagship AI systems, not internal tooling
Own AI system quality — build lightweight eval harnesses, monitor for hallucination or drift, and enforce guardrails against prompt injection and data leakage, especially on healthcare-adjacent and financial-data products
Build and maintain full-stack applications with clean, scalable code
Design and integrate databases, APIs, and third-party services
Automate repetitive dev and business workflows using AI agents
Collaborate with design and strategy teams to ship fast without cutting corners
Set up CI/CD pipelines, version control practices, and deployment workflows
Mentor and establish team coding standards, including how the team prompts, evaluates, and ships AI-assisted code
Required Technical Skills
Reworked around what an AI Engineer building production LLM systems actually needs day to day, plus the gaps specific to Worksmith's product surface (mobile, technical SEO, health data, and the health-check / CA workflow product line). AI/LLM engineering is now the lead skill set — full-stack ability is the delivery mechanism around it, not a separate track.
AI / LLM Engineering
LLM APIs & Tooling — Anthropic Claude API (Messages API, tool use / function calling, prompt caching, extended thinking); working familiarity with OpenAI or Gemini APIs a plus
Prompt & Context Engineering — system prompt design, few-shot patterning, context-window and token-budget management
Agentic Systems — building multi-step, tool-using agents (Claude Agent SDK, LangChain / LangGraph, CrewAI, or custom orchestration)
RAG & Vector Search — embeddings, chunking strategy, vector databases (Pinecone, Weaviate, pgvector, Chroma)
Structured Outputs & Tool Schemas — JSON Schema design, function-calling contracts, output validation and retry logic
AI Evaluation & Observability — eval harnesses, tracing and logging (LangSmith, Helicone, or custom scripts), regression testing for prompts
Model Selection & Cost/Latency Tradeoffs — knowing when to prompt vs. retrieve vs. fine-tune, and picking the right model tier per task
AI Security & Guardrails — prompt-injection defense, PII/PHI leakage prevention, least-privilege sandboxing of agent tool access
Financial & Operational Intelligence Systems
Python — primary language for AI integrations, automation, scripting, and data processing
JavaScript / TypeScript — primary language for frontend and backend
HTML5 / CSS3 — pixel-perfect UI implementation
SQL — relational database queries, schema design (PostgreSQL / MySQL)
Bash / Shell — server automation and DevOps scripting
Frontend
React.js or Next.js — SSR, dynamic routing, API routes
Tailwind CSS — or equivalent utility-first CSS
Responsive and mobile-first design implementation
Backend
Node.js / Express or FastAPI (Python) — FastAPI preferred where the service is AI-heavy
RESTful API design and GraphQL basics
Authentication systems (JWT, OAuth2)
Mobile (new — matches the Mobile Applications deliverable, previously unlisted)
React Native or Flutter — cross-platform mobile development
Push notification services (Firebase Cloud Messaging / APNs)
App store deployment and release management (Apple App Store, Google Play)
Databases & Storage
PostgreSQL / MySQL — relational data modeling
MongoDB or Firebase — NoSQL for flexible schemas
Redis — caching and session management
Vector stores — pgvector, Pinecone, Weaviate, or Chroma for embeddings and semantic search
DevOps & Deployment
Docker — containerization
GitHub Actions or similar CI/CD pipelines
Vercel / AWS / GCP — cloud deployment
Linux server management basics
Integrations & Automation
Nice to Have
A shipped AI agent or LLM-powered product in production, not just a prototype
Prior experience in a fast-paced agency or consulting environment juggling multiple client codebases
Comfort making architecture calls with incomplete specs and tight timelines