AI Developer
TaoScope
·
Engineering
$4k–$7k/month
Open
Remote
Full-time
Contract
AI
Build AI-powered features, model integrations, and intelligent automation that improve TaoScope for operators.
Apply
TaoScope is adding AI-assisted workflows to help operators investigate subnets, compare validators, and move faster through complex on-chain data. We need engineers who can integrate models responsibly — with evals, guardrails, observability, and honest failure modes.
You will ship production AI features and the automation around them: tool integrations, eval suites, cost controls, and runbooks. This role suits builders who care about measurable quality as much as demo velocity.
We want production AI engineering with evals and monitoring — not prompt-only experiments.
What you'll work on
LLM integrations for operator-facing summaries, triage, and internal tooling
Tool-using flows grounded in TaoScope APIs and chain-derived data
Eval suites, regression tests, and quality thresholds for AI features
Automation pipelines that keep datasets, prompts, and model outputs reliable
Observability for latency, cost, and failure modes in model-backed services
Collaboration with product and RAG engineers on retrieval contracts and citations
Documentation and runbooks so AI workflows are maintainable after handoff
Safe defaults around auth, rate limits, and sensitive data boundaries
Requirements
·
Shipped AI or LLM-backed features in production with monitoring and eval discipline
·
Strong TypeScript or Python with API integration experience
·
Comfort defining quality metrics and catching regressions before users do
·
Experience with structured outputs, tool calling, or workflow orchestration
·
Ownership mindset: you maintain what you build after the first merge
·
Examples of AI or automation work you can walk through — repos or write-ups
Nice to have
+
Agent frameworks, copilots, or multi-step planners in technical products
+
RAG pipelines, vector search, or retrieval quality metrics
+
Bittensor, on-chain analytics, or staking context
+
GitHub Actions, schedulers, or workflow automation at scale
We're not looking for
Prompt-only hobby projects with no monitoring or evals
Automation that silently swallows errors or retries forever without alerting
Demo builders who cannot explain latency, cost, or failure rates in production
Why join
·
Apply AI where ground truth exists — chain data and explorer APIs
·
Force multiplier on a lean team shipping operator-facing intelligence
·
Clear backlog of AI and automation wins waiting for an owner
·
Remote, async-friendly collaboration with engineers who value good tooling