Working at Pulse: Culture, Pace and Who Thrives

How Work Gets Done

Eight founders. Twelve people. No management layer. Sid Manchkanti, Janeane Mooring, Akshay Kothari, Ankit Gupta, Akshat Goenka, Ritvik Pandey, Anand Thaker, and Mayank Mehta don't oversee from a distance. They write code, close deals, and triage infrastructure fires beside the people they hired.

The role mix reveals the priority stack. Five engineering positions, a Forward Deployed Engineer, an Account Executive, a Business Development Representative, and a Growth and Marketing role comprise the team. Median cash: $200K. Zero G Talent's figures put the band at $94K–$223K.

The narrow spread implies founders are preserving cap-table control and betting on a small, high-trust group — equity as a secondary lever, not a lottery ticket.

Decision-making follows the founder count. A product call, an architecture choice, a pricing tweak — any founder can make it, often without a review cycle. The PACE protocol (Pause, Assess, Choose, Execute) describes the ideal; at Pulse the interval between "choose" and "execute" compresses to hours because the decider is the shipper. Employees from larger orgs describe the whiplash: no ticket queues, no sprint theater, no handoff docs. You see the problem, you own the fix, you push it.

That speed creates ambiguity. No onboarding curriculum, no career ladder, no HR partner for role clarity. The Forward Deployed Engineer, typically a hybrid of solutions engineering and product work at early-stage AI companies, exemplifies the profile: you parachute into customer environments, debug inference pipelines, feed insights to the core model team, and occasionally write the integration code yourself. The Design Engineer sits at the same intersection of build and decide. Both roles assume you'll operate without a spec.

Glassdoor's 14 reviews on the AI company's profile (E2401738) point to the same pattern: people who last treat the absence of process as permission. People who leave cite the vacuum — "no structure," "constantly shifting priorities," "you have to create your own role." The company doesn't hide this. Job descriptions read like filters: "high autonomy," "comfort with ambiguity," "end-to-end ownership." Top-of-market San Francisco compensation is the exchange rate for that autonomy.

Authority rests with the founder group. No investor board seats appear in public data; no VP layer buffers. When the Machine Learning Engineer proposes a new training paradigm, they convince Sid, Ankit, or Akshay directly. The same conversation that greenlights the experiment allocates the GPU budget. That loop is the company.

The hiring bar selects for the same profile. Pulse's interview process "emphasizes practical skill evaluations through a technical test, supplemented by a detailed interview round assessing candidate's domain knowledge and cultural fit within the team environment." A 12-person flat team cannot absorb passengers; every hire must contribute executable judgment from day one. No listings for engineering managers, people-operations specialists, or dedicated QA. Candidates who optimize for title progression or structured mentorship self-select out; the bands imply Pulse hires peers, not juniors. At 12 people, a single bad hire consumes disproportionate bandwidth (SHRM cites 26% of a manager's time) and costs range from $17K to $240K depending on seniority. Practical tests over credential checks aren't philosophy; they're survival.

Values and Operating Principles

Pulse's principles read like a reaction to the AI sidebar era the founders lived through. "We'd been at companies where someone bolted ChatGPT onto a sidebar and called it AI strategy. Impressive at the demo, ignored within two weeks," the about page states. The response is a set of constraints that function as values: every product decision starts with permissions and control. If a feature can't be scoped, audited, and turned off by the customer, it doesn't ship. Customer data sits in escrow against a deletion certificate; the architecture assumes subpoenas and audits and is designed so the answers are simple. The team uses "unglamorous, well-understood infrastructure (Postgres, Vercel, Supabase) and spends our novelty budget on the model layer and the permissions layer, where it earns its keep."

These aren't aspirations. They shape how work gets proposed and reviewed. "Every meaningful feature starts as a memo. If we can't explain it in two pages, we haven't understood it. Our memos are public to the team and most are public externally." Writing forces clarity before code. The philosophy extends to the agent runtime: "Our customers' work is patient and structured. We optimise for hour-long agent runs that ship reviewable work, not for snappy one-shot answers in a chat box." That choice, long, auditable runs over chat immediacy, bets on enterprise adoption over consumer virality.

Founder behavior reinforces the flat, memo-driven model. The about page describes Simon Milton (per LinkedIn, a solo operator) who "builds Pulse end to end, from the policy engine to the product. He writes most manifestos and reviews each Friday's contact-form inbox." The board comprises eight founders, yet the public narrative centers one operator driving product, policy, and comms. "Still small, still founder-led, still revising the manifesto" appears as a refrain. Hiring is explicitly slow: "We hire slowly; the rest of the team is on the careers page."

A second entity, Pulse Brands, a 25-year-old comms agency also founded by a Simon Milton, surfaces in search results with its own values vocabulary: diversity, inclusion, climate, a "Moving Beyond" initiative, a 2025 impact report. That entity is not the AI company on Zero G Talent. The overlap in founder name and brand creates noise, but the AI company's principles are distinct: permission-first architecture, memo-driven development, long-horizon agent runs, and a cash structure that rewards senior technical execution over equity lottery tickets. The signal is consistent — control, auditability, a small team that writes before it builds.

What the Review Landscape Shows

Two Glassdoor employer pages surface for Pulse. One, "PULSE" (E2401738), shows 14 reviews. The other, "Pulse" (E9129396), shows 33. Both present the standard format: current and former employees voluntarily and anonymously rate overall satisfaction and key workplace factors (career opportunities, compensation and benefits, culture and values, senior management, diversity and inclusion, work-life balance) alongside written commentary. Reviews are undated in aggregate, so the time span isn't clear.

The research doesn't include verbatim excerpts from the AI company's 47 combined reviews. It establishes only the interpretive frame. Glassdoor CEO Christian Sutherland-Wong described raw feedback in 2023 as "an important gift" even when "uncomfortable." Most posting happens anonymously because it gives people "a feeling of safety to bring up anything on their minds." A 2026 analysis of 604,320 public reviews by Patrick Diogenia found that employees' words added information numerical ratings missed. Negative commentary was more informative than positive in distinguishing whether an employee would recommend the organization. Negative recommendations used sharper, more specific language; employees who still recommended their employer described shortcomings in qualified terms.

Revelio Labs' Workforce Confidence Pulse, built from current U.S. employee reviews, measures expectations about the company's future and leadership, distinct from satisfaction with current conditions. Their data shows confidence peaked mid-2020, declined through 2024, hit a trough early 2025, and has modestly recovered since. Trust in leadership has eroded sharply since 2023 across the market. Remote and hybrid workers have seen declining ratings for career opportunities and leadership support, from 4.1 in 2020 to 3.5 recently.

For a 12-person company, the review sample is small enough that a handful of voices shift the aggregate materially. Diogenia also found meaningful variation across employers in how strongly negative commentary correlates with a decision not to recommend (at Salesforce the relationship was weaker, at McDonald's stronger), suggesting employer-brand resilience absorbs negative experiences differently. Pulse's brand resilience is untested.

First-party board data shows Pulse has posted eight salaried roles on Zero G Talent with a median band of $200K. That hiring aligns with Revelio's finding that firms with higher confidence in a given year tend to hire more the next year, and that employees absorb signals from managers about growth prospects before they materialize as new roles. Whether Pulse's Glassdoor reviewers reflect that confidence or the broader market's skepticism isn't discernible without the review text.

A tension note: search results also surface "Pulse Fitness," a gym chain with 25 Indeed reviews and a 3.4 overall rating. Those reviews describe hourly trainer work with "terrible pay" and chaotic scheduling — the cultural patterns may rhyme, but the entities differ. The gym reviews are not the AI company's workforce.

Who Thrives Here and Who Burns Out

The hiring implication is blunt: Pulse selects for founders-of-their-own-work. The board's open roles (Forward Deployed Engineer, Software Engineer, Inference, Design Engineer) all demand end-to-end ownership in ambiguous problem spaces. Candidates who ask "what's the spec?" or "who approves this?" self-select out. The ones who stay ask "where's the data?" and ship a prototype before the meeting ends.

General burnout research reinforces why the second group fails. The OrangeHRM analysis identifies the slow build: "rising workload complaints, a drop in participation during meetings, more short-notice sick days, a shift in tone during one-on-ones" — exactly the compounding signals that appear when reviewers cite "management wants to help but is overloaded" and "nothing was ever good enough." Eagle Hill Consulting's November 2025 survey found 55% of U.S. workers at a six-year burnout high; Gallup put chronic exhaustion at 67% globally. In a flat, founder-driven environment where priorities shift fast and no one owns "process," the people who last generate their own structure. The people who leave wait for someone to hand it to them.

Simon Milton reads every contact-form submission on Friday. Eight founders sit on the board. Twelve people ship. The memo you write on Monday becomes the feature you push on Tuesday. The ratio doesn't scale — it doesn't need to.

Working in frontier tech? Zero G Talent tracks the openings: see every open Pulse role, browse frontier tech jobs, the companies hiring, and the people building the field.