Harvey
About this role
Harvey builds AI for law firms, and its core infrastructure is what every user interaction runs
through: billions of prompt tokens and millions of daily requests across a global platform. This seat
designs and builds new infrastructure systems while scaling and hardening what already exists,
balancing new construction against operational excellence as the product adds regions, customers and
load. Concrete examples from the posting: architecting distributed systems for reliability with load
balancing, quota management and failover, and building distributed rate limiting and quota systems on
Redis backed algorithms to absorb bursty traffic without degrading the experience. Observability is
Datadog and Sentry with PagerDuty for incidents. Apply if you have run infrastructure for a product
where downtime is visible to demanding enterprise customers.
through: billions of prompt tokens and millions of daily requests across a global platform. This seat
designs and builds new infrastructure systems while scaling and hardening what already exists,
balancing new construction against operational excellence as the product adds regions, customers and
load. Concrete examples from the posting: architecting distributed systems for reliability with load
balancing, quota management and failover, and building distributed rate limiting and quota systems on
Redis backed algorithms to absorb bursty traffic without degrading the experience. Observability is
Datadog and Sentry with PagerDuty for incidents. Apply if you have run infrastructure for a product
where downtime is visible to demanding enterprise customers.
Who this is for
Requirements. 10+ years of experience in infrastructure engineering or platform engineering in a
production environment. Experience with observability tools including Datadog and Sentry, and incident
response practices including PagerDuty.
What you would do. Design and build new infrastructure systems while scaling and strengthening
existing infrastructure. Architect and optimise distributed systems for reliability, including load
balancing, quota management and failover mechanisms. Build distributed rate limiting and quota
management systems using Redis backed algorithms to handle bursty traffic patterns without degrading
user experience. Work in an environment balanced between building new systems and operational
excellence, keeping Harvey resilient and efficient as it scales across products, regions, customers
and usage. Your work affects the reliability, scalability and security of the platform serving
leading law firms and professional service providers.
Scale, as stated. Harvey's infrastructure processes billions of prompt tokens and millions of daily
requests across its global legal AI platform.
⛔ Location, because the job board is wrong. This posting is flagged remote on Harvey's own board, and
its body says "This role is based in Bengaluru, India". Every Harvey India posting carries the same
false remote flag, and one Harvey posting listed under Bengaluru states in its body that it is based
in San Francisco. We take the location from the body every time. This is a Bengaluru role.
Eligibility. Harvey states on its India postings that you must be authorised to work in India and that
visa sponsorship is not available.
Who this is for. An infrastructure or platform engineer with ten or more years who wants
high throughput AI serving problems. Rate limiting, quota management and failover at this scale are
specific skills; if you have built them, the posting is describing your work directly. The company is
explicit about pace and intensity in its culture section, which is worth reading honestly before
applying.
production environment. Experience with observability tools including Datadog and Sentry, and incident
response practices including PagerDuty.
What you would do. Design and build new infrastructure systems while scaling and strengthening
existing infrastructure. Architect and optimise distributed systems for reliability, including load
balancing, quota management and failover mechanisms. Build distributed rate limiting and quota
management systems using Redis backed algorithms to handle bursty traffic patterns without degrading
user experience. Work in an environment balanced between building new systems and operational
excellence, keeping Harvey resilient and efficient as it scales across products, regions, customers
and usage. Your work affects the reliability, scalability and security of the platform serving
leading law firms and professional service providers.
Scale, as stated. Harvey's infrastructure processes billions of prompt tokens and millions of daily
requests across its global legal AI platform.
⛔ Location, because the job board is wrong. This posting is flagged remote on Harvey's own board, and
its body says "This role is based in Bengaluru, India". Every Harvey India posting carries the same
false remote flag, and one Harvey posting listed under Bengaluru states in its body that it is based
in San Francisco. We take the location from the body every time. This is a Bengaluru role.
Eligibility. Harvey states on its India postings that you must be authorised to work in India and that
visa sponsorship is not available.
Who this is for. An infrastructure or platform engineer with ten or more years who wants
high throughput AI serving problems. Rate limiting, quota management and failover at this scale are
specific skills; if you have built them, the posting is describing your work directly. The company is
explicit about pace and intensity in its culture section, which is worth reading honestly before
applying.
Apply on company site
Opens jobs.ashbyhq.com, the employer's own application page. Applying is always free.